Mercurial > libavcodec.hg
changeset 2979:bfabfdf9ce55 libavcodec
COSMETICS: tabs --> spaces, some prettyprinting
line wrap: on
line diff
--- a/8bps.c Wed Dec 21 17:50:40 2005 +0000 +++ b/8bps.c Thu Dec 22 01:10:11 2005 +0000 @@ -44,11 +44,11 @@ */ typedef struct EightBpsContext { - AVCodecContext *avctx; - AVFrame pic; + AVCodecContext *avctx; + AVFrame pic; - unsigned char planes; - unsigned char planemap[4]; + unsigned char planes; + unsigned char planemap[4]; } EightBpsContext; @@ -59,87 +59,87 @@ */ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8_t *buf, int buf_size) { - EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; - unsigned char *encoded = (unsigned char *)buf; - unsigned char *pixptr, *pixptr_end; - unsigned int height = avctx->height; // Real image height - unsigned int dlen, p, row; - unsigned char *lp, *dp; - unsigned char count; - unsigned int px_inc; - unsigned int planes = c->planes; - unsigned char *planemap = c->planemap; + EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; + unsigned char *encoded = (unsigned char *)buf; + unsigned char *pixptr, *pixptr_end; + unsigned int height = avctx->height; // Real image height + unsigned int dlen, p, row; + unsigned char *lp, *dp; + unsigned char count; + unsigned int px_inc; + unsigned int planes = c->planes; + unsigned char *planemap = c->planemap; - if(c->pic.data[0]) - avctx->release_buffer(avctx, &c->pic); + if(c->pic.data[0]) + avctx->release_buffer(avctx, &c->pic); - c->pic.reference = 0; - c->pic.buffer_hints = FF_BUFFER_HINTS_VALID; - if(avctx->get_buffer(avctx, &c->pic) < 0){ - av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); - return -1; - } + c->pic.reference = 0; + c->pic.buffer_hints = FF_BUFFER_HINTS_VALID; + if(avctx->get_buffer(avctx, &c->pic) < 0){ + av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); + return -1; + } - /* Set data pointer after line lengths */ - dp = encoded + planes * (height << 1); + /* Set data pointer after line lengths */ + dp = encoded + planes * (height << 1); - /* Ignore alpha plane, don't know what to do with it */ - if (planes == 4) - planes--; + /* Ignore alpha plane, don't know what to do with it */ + if (planes == 4) + planes--; - px_inc = planes + (avctx->pix_fmt == PIX_FMT_RGBA32); + px_inc = planes + (avctx->pix_fmt == PIX_FMT_RGBA32); - for (p = 0; p < planes; p++) { - /* Lines length pointer for this plane */ - lp = encoded + p * (height << 1); + for (p = 0; p < planes; p++) { + /* Lines length pointer for this plane */ + lp = encoded + p * (height << 1); - /* Decode a plane */ - for(row = 0; row < height; row++) { - pixptr = c->pic.data[0] + row * c->pic.linesize[0] + planemap[p]; - pixptr_end = pixptr + c->pic.linesize[0]; - dlen = be2me_16(*(unsigned short *)(lp+row*2)); - /* Decode a row of this plane */ - while(dlen > 0) { - if(dp + 1 >= buf+buf_size) return -1; - if ((count = *dp++) <= 127) { - count++; - dlen -= count + 1; - if (pixptr + count * px_inc > pixptr_end) - break; - if(dp + count > buf+buf_size) return -1; - while(count--) { - *pixptr = *dp++; - pixptr += px_inc; - } - } else { - count = 257 - count; - if (pixptr + count * px_inc > pixptr_end) - break; - while(count--) { - *pixptr = *dp; - pixptr += px_inc; - } - dp++; - dlen -= 2; - } - } - } - } + /* Decode a plane */ + for(row = 0; row < height; row++) { + pixptr = c->pic.data[0] + row * c->pic.linesize[0] + planemap[p]; + pixptr_end = pixptr + c->pic.linesize[0]; + dlen = be2me_16(*(unsigned short *)(lp+row*2)); + /* Decode a row of this plane */ + while(dlen > 0) { + if(dp + 1 >= buf+buf_size) return -1; + if ((count = *dp++) <= 127) { + count++; + dlen -= count + 1; + if (pixptr + count * px_inc > pixptr_end) + break; + if(dp + count > buf+buf_size) return -1; + while(count--) { + *pixptr = *dp++; + pixptr += px_inc; + } + } else { + count = 257 - count; + if (pixptr + count * px_inc > pixptr_end) + break; + while(count--) { + *pixptr = *dp; + pixptr += px_inc; + } + dp++; + dlen -= 2; + } + } + } + } - if (avctx->palctrl) { - memcpy (c->pic.data[1], avctx->palctrl->palette, AVPALETTE_SIZE); - if (avctx->palctrl->palette_changed) { - c->pic.palette_has_changed = 1; - avctx->palctrl->palette_changed = 0; - } else - c->pic.palette_has_changed = 0; - } + if (avctx->palctrl) { + memcpy (c->pic.data[1], avctx->palctrl->palette, AVPALETTE_SIZE); + if (avctx->palctrl->palette_changed) { + c->pic.palette_has_changed = 1; + avctx->palctrl->palette_changed = 0; + } else + c->pic.palette_has_changed = 0; + } - *data_size = sizeof(AVFrame); - *(AVFrame*)data = c->pic; + *data_size = sizeof(AVFrame); + *(AVFrame*)data = c->pic; - /* always report that the buffer was completely consumed */ - return buf_size; + /* always report that the buffer was completely consumed */ + return buf_size; } @@ -150,53 +150,53 @@ */ static int decode_init(AVCodecContext *avctx) { - EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; + EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; - c->avctx = avctx; - avctx->has_b_frames = 0; + c->avctx = avctx; + avctx->has_b_frames = 0; - c->pic.data[0] = NULL; + c->pic.data[0] = NULL; if (avcodec_check_dimensions(avctx, avctx->width, avctx->height) < 0) { return 1; } - switch (avctx->bits_per_sample) { - case 8: - avctx->pix_fmt = PIX_FMT_PAL8; - c->planes = 1; - c->planemap[0] = 0; // 1st plane is palette indexes - if (avctx->palctrl == NULL) { - av_log(avctx, AV_LOG_ERROR, "Error: PAL8 format but no palette from demuxer.\n"); - return -1; - } - break; - case 24: - avctx->pix_fmt = avctx->get_format(avctx, pixfmt_rgb24); - c->planes = 3; - c->planemap[0] = 2; // 1st plane is red - c->planemap[1] = 1; // 2nd plane is green - c->planemap[2] = 0; // 3rd plane is blue - break; - case 32: - avctx->pix_fmt = PIX_FMT_RGBA32; - c->planes = 4; + switch (avctx->bits_per_sample) { + case 8: + avctx->pix_fmt = PIX_FMT_PAL8; + c->planes = 1; + c->planemap[0] = 0; // 1st plane is palette indexes + if (avctx->palctrl == NULL) { + av_log(avctx, AV_LOG_ERROR, "Error: PAL8 format but no palette from demuxer.\n"); + return -1; + } + break; + case 24: + avctx->pix_fmt = avctx->get_format(avctx, pixfmt_rgb24); + c->planes = 3; + c->planemap[0] = 2; // 1st plane is red + c->planemap[1] = 1; // 2nd plane is green + c->planemap[2] = 0; // 3rd plane is blue + break; + case 32: + avctx->pix_fmt = PIX_FMT_RGBA32; + c->planes = 4; #ifdef WORDS_BIGENDIAN - c->planemap[0] = 1; // 1st plane is red - c->planemap[1] = 2; // 2nd plane is green - c->planemap[2] = 3; // 3rd plane is blue - c->planemap[3] = 0; // 4th plane is alpha??? + c->planemap[0] = 1; // 1st plane is red + c->planemap[1] = 2; // 2nd plane is green + c->planemap[2] = 3; // 3rd plane is blue + c->planemap[3] = 0; // 4th plane is alpha??? #else - c->planemap[0] = 2; // 1st plane is red - c->planemap[1] = 1; // 2nd plane is green - c->planemap[2] = 0; // 3rd plane is blue - c->planemap[3] = 3; // 4th plane is alpha??? + c->planemap[0] = 2; // 1st plane is red + c->planemap[1] = 1; // 2nd plane is green + c->planemap[2] = 0; // 3rd plane is blue + c->planemap[3] = 3; // 4th plane is alpha??? #endif - break; - default: - av_log(avctx, AV_LOG_ERROR, "Error: Unsupported color depth: %u.\n", avctx->bits_per_sample); - return -1; - } + break; + default: + av_log(avctx, AV_LOG_ERROR, "Error: Unsupported color depth: %u.\n", avctx->bits_per_sample); + return -1; + } return 0; } @@ -211,24 +211,24 @@ */ static int decode_end(AVCodecContext *avctx) { - EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; + EightBpsContext * const c = (EightBpsContext *)avctx->priv_data; - if (c->pic.data[0]) - avctx->release_buffer(avctx, &c->pic); + if (c->pic.data[0]) + avctx->release_buffer(avctx, &c->pic); - return 0; + return 0; } AVCodec eightbps_decoder = { - "8bps", - CODEC_TYPE_VIDEO, - CODEC_ID_8BPS, - sizeof(EightBpsContext), - decode_init, - NULL, - decode_end, - decode_frame, - CODEC_CAP_DR1, + "8bps", + CODEC_TYPE_VIDEO, + CODEC_ID_8BPS, + sizeof(EightBpsContext), + decode_init, + NULL, + decode_end, + decode_frame, + CODEC_CAP_DR1, };
--- a/a52dec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/a52dec.c Thu Dec 22 01:10:11 2005 +0000 @@ -58,11 +58,11 @@ a52_state_t* (*a52_init)(uint32_t mm_accel); sample_t* (*a52_samples)(a52_state_t * state); int (*a52_syncinfo)(uint8_t * buf, int * flags, - int * sample_rate, int * bit_rate); + int * sample_rate, int * bit_rate); int (*a52_frame)(a52_state_t * state, uint8_t * buf, int * flags, - sample_t * level, sample_t bias); + sample_t * level, sample_t bias); void (*a52_dynrng)(a52_state_t * state, - sample_t (* call) (sample_t, void *), void * data); + sample_t (* call) (sample_t, void *), void * data); int (*a52_block)(a52_state_t * state); void (*a52_free)(a52_state_t * state); @@ -105,7 +105,7 @@ if (!s->a52_init || !s->a52_samples || !s->a52_syncinfo || !s->a52_frame || !s->a52_block || !s->a52_free) { - dlclose(s->handle); + dlclose(s->handle); return -1; } #else @@ -130,22 +130,22 @@ static inline int blah (int32_t i) { if (i > 0x43c07fff) - return 32767; + return 32767; else if (i < 0x43bf8000) - return -32768; + return -32768; return i - 0x43c00000; } static inline void float_to_int (float * _f, int16_t * s16, int nchannels) { int i, j, c; - int32_t * f = (int32_t *) _f; // XXX assumes IEEE float format + int32_t * f = (int32_t *) _f; // XXX assumes IEEE float format j = 0; nchannels *= 256; for (i = 0; i < 256; i++) { - for (c = 0; c < nchannels; c += 256) - s16[j++] = blah (f[i + c]); + for (c = 0; c < nchannels; c += 256) + s16[j++] = blah (f[i + c]); } } @@ -164,7 +164,7 @@ short *out_samples = data; float level; static const int ac3_channels[8] = { - 2, 1, 2, 3, 3, 4, 4, 5 + 2, 1, 2, 3, 3, 4, 4, 5 }; buf_ptr = buf; @@ -186,20 +186,20 @@ memcpy(s->inbuf, s->inbuf + 1, HEADER_SIZE - 1); s->inbuf_ptr--; } else { - s->frame_size = len; + s->frame_size = len; /* update codec info */ avctx->sample_rate = sample_rate; s->channels = ac3_channels[s->flags & 7]; if (s->flags & A52_LFE) - s->channels++; - if (avctx->channels == 0) - /* No specific number of channel requested */ - avctx->channels = s->channels; - else if (s->channels < avctx->channels) { - av_log(avctx, AV_LOG_ERROR, "ac3dec: AC3 Source channels are less than specified: output to %d channels.. (frmsize: %d)\n", s->channels, len); - avctx->channels = s->channels; - } - avctx->bit_rate = bit_rate; + s->channels++; + if (avctx->channels == 0) + /* No specific number of channel requested */ + avctx->channels = s->channels; + else if (s->channels < avctx->channels) { + av_log(avctx, AV_LOG_ERROR, "ac3dec: AC3 Source channels are less than specified: output to %d channels.. (frmsize: %d)\n", s->channels, len); + avctx->channels = s->channels; + } + avctx->bit_rate = bit_rate; } } } else if (len < s->frame_size) {
--- a/ac3dec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ac3dec.c Thu Dec 22 01:10:11 2005 +0000 @@ -54,23 +54,23 @@ static inline int blah (int32_t i) { if (i > 0x43c07fff) - return 32767; + return 32767; else if (i < 0x43bf8000) - return -32768; + return -32768; else - return i - 0x43c00000; + return i - 0x43c00000; } static inline void float_to_int (float * _f, int16_t * s16, int nchannels) { int i, j, c; - int32_t * f = (int32_t *) _f; // XXX assumes IEEE float format + int32_t * f = (int32_t *) _f; // XXX assumes IEEE float format j = 0; nchannels *= 256; for (i = 0; i < 256; i++) { - for (c = 0; c < nchannels; c += 256) - s16[j++] = blah (f[i + c]); + for (c = 0; c < nchannels; c += 256) + s16[j++] = blah (f[i + c]); } } @@ -89,7 +89,7 @@ short *out_samples = data; float level; static const int ac3_channels[8] = { - 2, 1, 2, 3, 3, 4, 4, 5 + 2, 1, 2, 3, 3, 4, 4, 5 }; buf_ptr = buf; @@ -111,20 +111,20 @@ memcpy(s->inbuf, s->inbuf + 1, HEADER_SIZE - 1); s->inbuf_ptr--; } else { - s->frame_size = len; + s->frame_size = len; /* update codec info */ avctx->sample_rate = sample_rate; s->channels = ac3_channels[s->flags & 7]; if (s->flags & AC3_LFE) - s->channels++; - if (avctx->channels == 0) - /* No specific number of channel requested */ - avctx->channels = s->channels; - else if (s->channels < avctx->channels) { + s->channels++; + if (avctx->channels == 0) + /* No specific number of channel requested */ + avctx->channels = s->channels; + else if (s->channels < avctx->channels) { av_log( avctx, AV_LOG_INFO, "ac3dec: AC3 Source channels are less than specified: output to %d channels.. (frmsize: %d)\n", s->channels, len); - avctx->channels = s->channels; - } - avctx->bit_rate = bit_rate; + avctx->channels = s->channels; + } + avctx->bit_rate = bit_rate; } } } else if (len < s->frame_size) {
--- a/ac3enc.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ac3enc.c Thu Dec 22 01:10:11 2005 +0000 @@ -337,8 +337,8 @@ /* do a 2^n point complex fft on 2^ln points. */ static void fft(IComplex *z, int ln) { - int j, l, np, np2; - int nblocks, nloops; + int j, l, np, np2; + int nblocks, nloops; register IComplex *p,*q; int tmp_re, tmp_im; @@ -472,7 +472,7 @@ exp_strategy[i][ch] = EXP_REUSE; } if (is_lfe) - return; + return; /* now select the encoding strategy type : if exponents are often recoded, we use a coarse encoding */ @@ -493,7 +493,7 @@ exp_strategy[i][ch] = EXP_D15; break; } - i = j; + i = j; } } @@ -553,9 +553,9 @@ /* Decrease the delta between each groups to within 2 * so that they can be differentially encoded */ for (i=1;i<=nb_groups;i++) - exp1[i] = FFMIN(exp1[i], exp1[i-1] + 2); + exp1[i] = FFMIN(exp1[i], exp1[i-1] + 2); for (i=nb_groups-1;i>=0;i--) - exp1[i] = FFMIN(exp1[i], exp1[i+1] + 2); + exp1[i] = FFMIN(exp1[i], exp1[i+1] + 2); /* now we have the exponent values the decoder will see */ encoded_exp[0] = exp1[0]; @@ -708,8 +708,8 @@ if(i==0) frame_bits += 4; } frame_bits += 2 * s->nb_channels; /* chexpstr[2] * c */ - if (s->lfe) - frame_bits++; /* lfeexpstr */ + if (s->lfe) + frame_bits++; /* lfeexpstr */ for(ch=0;ch<s->nb_channels;ch++) { if (exp_strategy[i][ch] != EXP_REUSE) frame_bits += 6 + 2; /* chbwcod[6], gainrng[2] */ @@ -736,11 +736,11 @@ csnroffst = s->csnroffst; while (csnroffst >= 0 && - bit_alloc(s, bap, encoded_exp, exp_strategy, frame_bits, csnroffst, 0) < 0) - csnroffst -= SNR_INC1; + bit_alloc(s, bap, encoded_exp, exp_strategy, frame_bits, csnroffst, 0) < 0) + csnroffst -= SNR_INC1; if (csnroffst < 0) { - av_log(NULL, AV_LOG_ERROR, "Yack, Error !!!\n"); - return -1; + av_log(NULL, AV_LOG_ERROR, "Yack, Error !!!\n"); + return -1; } while ((csnroffst + SNR_INC1) <= 63 && bit_alloc(s, bap1, encoded_exp, exp_strategy, frame_bits, @@ -815,19 +815,19 @@ int i, j, ch; float alpha; static const uint8_t acmod_defs[6] = { - 0x01, /* C */ - 0x02, /* L R */ - 0x03, /* L C R */ - 0x06, /* L R SL SR */ - 0x07, /* L C R SL SR */ - 0x07, /* L C R SL SR (+LFE) */ + 0x01, /* C */ + 0x02, /* L R */ + 0x03, /* L C R */ + 0x06, /* L R SL SR */ + 0x07, /* L C R SL SR */ + 0x07, /* L C R SL SR (+LFE) */ }; avctx->frame_size = AC3_FRAME_SIZE; /* number of channels */ if (channels < 1 || channels > 6) - return -1; + return -1; s->acmod = acmod_defs[channels - 1]; s->lfe = (channels == 6) ? 1 : 0; s->nb_all_channels = channels; @@ -871,7 +871,7 @@ s->nb_coefs[ch] = ((s->chbwcod[ch] + 12) * 3) + 37; } if (s->lfe) { - s->nb_coefs[s->lfe_channel] = 7; /* fixed */ + s->nb_coefs[s->lfe_channel] = 7; /* fixed */ } /* initial snr offset */ s->csnroffst = 40; @@ -907,9 +907,9 @@ put_bits(&s->pb, 3, s->bsmod); put_bits(&s->pb, 3, s->acmod); if ((s->acmod & 0x01) && s->acmod != 0x01) - put_bits(&s->pb, 2, 1); /* XXX -4.5 dB */ + put_bits(&s->pb, 2, 1); /* XXX -4.5 dB */ if (s->acmod & 0x04) - put_bits(&s->pb, 2, 1); /* XXX -6 dB */ + put_bits(&s->pb, 2, 1); /* XXX -6 dB */ if (s->acmod == 0x02) put_bits(&s->pb, 2, 0); /* surround not indicated */ put_bits(&s->pb, 1, s->lfe); /* LFE */ @@ -995,20 +995,20 @@ if (s->acmod == 2) { - if(block_num==0) - { - /* first block must define rematrixing (rematstr) */ - put_bits(&s->pb, 1, 1); + if(block_num==0) + { + /* first block must define rematrixing (rematstr) */ + put_bits(&s->pb, 1, 1); - /* dummy rematrixing rematflg(1:4)=0 */ - for (rbnd=0;rbnd<4;rbnd++) - put_bits(&s->pb, 1, 0); - } - else - { - /* no matrixing (but should be used in the future) */ - put_bits(&s->pb, 1, 0); - } + /* dummy rematrixing rematflg(1:4)=0 */ + for (rbnd=0;rbnd<4;rbnd++) + put_bits(&s->pb, 1, 0); + } + else + { + /* no matrixing (but should be used in the future) */ + put_bits(&s->pb, 1, 0); + } } #if defined(DEBUG) @@ -1023,7 +1023,7 @@ } if (s->lfe) { - put_bits(&s->pb, 1, exp_strategy[s->lfe_channel]); + put_bits(&s->pb, 1, exp_strategy[s->lfe_channel]); } for(ch=0;ch<s->nb_channels;ch++) { @@ -1047,7 +1047,7 @@ group_size = 4; break; } - nb_groups = (s->nb_coefs[ch] + (group_size * 3) - 4) / (3 * group_size); + nb_groups = (s->nb_coefs[ch] + (group_size * 3) - 4) / (3 * group_size); p = encoded_exp[ch]; /* first exponent */ @@ -1075,8 +1075,8 @@ put_bits(&s->pb, 7, ((delta0 * 5 + delta1) * 5) + delta2); } - if (ch != s->lfe_channel) - put_bits(&s->pb, 2, 0); /* no gain range info */ + if (ch != s->lfe_channel) + put_bits(&s->pb, 2, 0); /* no gain range info */ } /* bit allocation info */
--- a/adpcm.c Wed Dec 21 17:50:40 2005 +0000 +++ b/adpcm.c Thu Dec 22 01:10:11 2005 +0000 @@ -300,7 +300,7 @@ } static int adpcm_encode_frame(AVCodecContext *avctx, - unsigned char *frame, int buf_size, void *data) + unsigned char *frame, int buf_size, void *data) { int n, i, st; short *samples; @@ -431,8 +431,8 @@ switch(avctx->codec->id) { case CODEC_ID_ADPCM_CT: - c->status[0].step = c->status[1].step = 511; - break; + c->status[0].step = c->status[1].step = 511; + break; default: break; } @@ -498,16 +498,16 @@ predictor = c->predictor; /* predictor update is not so trivial: predictor is multiplied on 254/256 before updating */ if(sign) - predictor = ((predictor * 254) >> 8) - diff; + predictor = ((predictor * 254) >> 8) - diff; else - predictor = ((predictor * 254) >> 8) + diff; + predictor = ((predictor * 254) >> 8) + diff; /* calculate new step and clamp it to range 511..32767 */ new_step = (ct_adpcm_table[nibble & 7] * c->step) >> 8; c->step = new_step; if(c->step < 511) - c->step = 511; + c->step = 511; if(c->step > 32767) - c->step = 32767; + c->step = 32767; CLAMP_TO_SHORT(predictor); c->predictor = predictor; @@ -612,8 +612,8 @@ } static int adpcm_decode_frame(AVCodecContext *avctx, - void *data, int *data_size, - uint8_t *buf, int buf_size) + void *data, int *data_size, + uint8_t *buf, int buf_size) { ADPCMContext *c = avctx->priv_data; ADPCMChannelStatus *cs; @@ -701,7 +701,7 @@ cs->predictor -= 0x10000; CLAMP_TO_SHORT(cs->predictor); - // XXX: is this correct ??: *samples++ = cs->predictor; + // XXX: is this correct ??: *samples++ = cs->predictor; cs->step_index = *src++; if (cs->step_index < 0) cs->step_index = 0; @@ -710,19 +710,19 @@ } for(m=4; src < (buf + buf_size);) { - *samples++ = adpcm_ima_expand_nibble(&c->status[0], src[0] & 0x0F, 3); + *samples++ = adpcm_ima_expand_nibble(&c->status[0], src[0] & 0x0F, 3); if (st) *samples++ = adpcm_ima_expand_nibble(&c->status[1], src[4] & 0x0F, 3); *samples++ = adpcm_ima_expand_nibble(&c->status[0], (src[0] >> 4) & 0x0F, 3); - if (st) { + if (st) { *samples++ = adpcm_ima_expand_nibble(&c->status[1], (src[4] >> 4) & 0x0F, 3); - if (!--m) { - m=4; - src+=4; - } - } - src++; - } + if (!--m) { + m=4; + src+=4; + } + } + src++; + } break; case CODEC_ID_ADPCM_4XM: cs = &(c->status[0]); @@ -739,13 +739,13 @@ m= (buf_size - (src - buf))>>st; for(i=0; i<m; i++) { - *samples++ = adpcm_ima_expand_nibble(&c->status[0], src[i] & 0x0F, 4); + *samples++ = adpcm_ima_expand_nibble(&c->status[0], src[i] & 0x0F, 4); if (st) *samples++ = adpcm_ima_expand_nibble(&c->status[1], src[i+m] & 0x0F, 4); *samples++ = adpcm_ima_expand_nibble(&c->status[0], src[i] >> 4, 4); - if (st) + if (st) *samples++ = adpcm_ima_expand_nibble(&c->status[1], src[i+m] >> 4, 4); - } + } src += m<<st; @@ -958,7 +958,7 @@ } break; case CODEC_ID_ADPCM_CT: - while (src < buf + buf_size) { + while (src < buf + buf_size) { if (st) { *samples++ = adpcm_ct_expand_nibble(&c->status[0], (src[0] >> 4) & 0x0F); @@ -970,78 +970,78 @@ *samples++ = adpcm_ct_expand_nibble(&c->status[0], src[0] & 0x0F); } - src++; + src++; } break; case CODEC_ID_ADPCM_SWF: { - GetBitContext gb; - const int *table; - int k0, signmask; - int size = buf_size*8; + GetBitContext gb; + const int *table; + int k0, signmask; + int size = buf_size*8; - init_get_bits(&gb, buf, size); + init_get_bits(&gb, buf, size); - // first frame, read bits & inital values - if (!c->nb_bits) - { - c->nb_bits = get_bits(&gb, 2)+2; -// av_log(NULL,AV_LOG_INFO,"nb_bits: %d\n", c->nb_bits); - } + // first frame, read bits & inital values + if (!c->nb_bits) + { + c->nb_bits = get_bits(&gb, 2)+2; +// av_log(NULL,AV_LOG_INFO,"nb_bits: %d\n", c->nb_bits); + } - table = swf_index_tables[c->nb_bits-2]; - k0 = 1 << (c->nb_bits-2); - signmask = 1 << (c->nb_bits-1); + table = swf_index_tables[c->nb_bits-2]; + k0 = 1 << (c->nb_bits-2); + signmask = 1 << (c->nb_bits-1); - while (get_bits_count(&gb) <= size) - { - int i; + while (get_bits_count(&gb) <= size) + { + int i; - c->nb_samples++; - // wrap around at every 4096 samples... - if ((c->nb_samples & 0xfff) == 1) - { - for (i = 0; i <= st; i++) - { - *samples++ = c->status[i].predictor = get_sbits(&gb, 16); - c->status[i].step_index = get_bits(&gb, 6); - } - } + c->nb_samples++; + // wrap around at every 4096 samples... + if ((c->nb_samples & 0xfff) == 1) + { + for (i = 0; i <= st; i++) + { + *samples++ = c->status[i].predictor = get_sbits(&gb, 16); + c->status[i].step_index = get_bits(&gb, 6); + } + } - // similar to IMA adpcm - for (i = 0; i <= st; i++) - { - int delta = get_bits(&gb, c->nb_bits); - int step = step_table[c->status[i].step_index]; - long vpdiff = 0; // vpdiff = (delta+0.5)*step/4 - int k = k0; + // similar to IMA adpcm + for (i = 0; i <= st; i++) + { + int delta = get_bits(&gb, c->nb_bits); + int step = step_table[c->status[i].step_index]; + long vpdiff = 0; // vpdiff = (delta+0.5)*step/4 + int k = k0; - do { - if (delta & k) - vpdiff += step; - step >>= 1; - k >>= 1; - } while(k); - vpdiff += step; + do { + if (delta & k) + vpdiff += step; + step >>= 1; + k >>= 1; + } while(k); + vpdiff += step; - if (delta & signmask) - c->status[i].predictor -= vpdiff; - else - c->status[i].predictor += vpdiff; + if (delta & signmask) + c->status[i].predictor -= vpdiff; + else + c->status[i].predictor += vpdiff; - c->status[i].step_index += table[delta & (~signmask)]; + c->status[i].step_index += table[delta & (~signmask)]; - c->status[i].step_index = clip(c->status[i].step_index, 0, 88); - c->status[i].predictor = clip(c->status[i].predictor, -32768, 32767); + c->status[i].step_index = clip(c->status[i].step_index, 0, 88); + c->status[i].predictor = clip(c->status[i].predictor, -32768, 32767); - *samples++ = c->status[i].predictor; - } - } + *samples++ = c->status[i].predictor; + } + } -// src += get_bits_count(&gb)*8; - src += size; +// src += get_bits_count(&gb)*8; + src += size; - break; + break; } case CODEC_ID_ADPCM_YAMAHA: while (src < buf + buf_size) {
--- a/allcodecs.c Wed Dec 21 17:50:40 2005 +0000 +++ b/allcodecs.c Thu Dec 22 01:10:11 2005 +0000 @@ -35,7 +35,7 @@ static int inited = 0; if (inited != 0) - return; + return; inited = 1; /* encoders */
--- a/alpha/asm.h Wed Dec 21 17:50:40 2005 +0000 +++ b/alpha/asm.h Thu Dec 22 01:10:11 2005 +0000 @@ -84,24 +84,24 @@ } *) (p))->__l) = l; \ } while (0) struct unaligned_long { uint64_t l; } __attribute__((packed)); -#define ldq_u(p) (*(const uint64_t *) (((uint64_t) (p)) & ~7ul)) -#define uldq(a) (((const struct unaligned_long *) (a))->l) +#define ldq_u(p) (*(const uint64_t *) (((uint64_t) (p)) & ~7ul)) +#define uldq(a) (((const struct unaligned_long *) (a))->l) #if GNUC_PREREQ(3,3) #define prefetch(p) __builtin_prefetch((p), 0, 1) #define prefetch_en(p) __builtin_prefetch((p), 0, 0) #define prefetch_m(p) __builtin_prefetch((p), 1, 1) #define prefetch_men(p) __builtin_prefetch((p), 1, 0) -#define cmpbge __builtin_alpha_cmpbge +#define cmpbge __builtin_alpha_cmpbge /* Avoid warnings. */ -#define extql(a, b) __builtin_alpha_extql(a, (uint64_t) (b)) -#define extwl(a, b) __builtin_alpha_extwl(a, (uint64_t) (b)) -#define extqh(a, b) __builtin_alpha_extqh(a, (uint64_t) (b)) -#define zap __builtin_alpha_zap -#define zapnot __builtin_alpha_zapnot -#define amask __builtin_alpha_amask -#define implver __builtin_alpha_implver -#define rpcc __builtin_alpha_rpcc +#define extql(a, b) __builtin_alpha_extql(a, (uint64_t) (b)) +#define extwl(a, b) __builtin_alpha_extwl(a, (uint64_t) (b)) +#define extqh(a, b) __builtin_alpha_extqh(a, (uint64_t) (b)) +#define zap __builtin_alpha_zap +#define zapnot __builtin_alpha_zapnot +#define amask __builtin_alpha_amask +#define implver __builtin_alpha_implver +#define rpcc __builtin_alpha_rpcc #else #define prefetch(p) asm volatile("ldl $31,%0" : : "m"(*(const char *) (p)) : "memory") #define prefetch_en(p) asm volatile("ldq $31,%0" : : "m"(*(const char *) (p)) : "memory") @@ -113,26 +113,26 @@ #define extqh(a, b) ({ uint64_t __r; asm ("extqh %r1,%2,%0" : "=r" (__r) : "rJ" (a), "rI" (b)); __r; }) #define zap(a, b) ({ uint64_t __r; asm ("zap %r1,%2,%0" : "=r" (__r) : "rJ" (a), "rI" (b)); __r; }) #define zapnot(a, b) ({ uint64_t __r; asm ("zapnot %r1,%2,%0" : "=r" (__r) : "rJ" (a), "rI" (b)); __r; }) -#define amask(a) ({ uint64_t __r; asm ("amask %1,%0" : "=r" (__r) : "rI" (a)); __r; }) -#define implver() ({ uint64_t __r; asm ("implver %0" : "=r" (__r)); __r; }) -#define rpcc() ({ uint64_t __r; asm volatile ("rpcc %0" : "=r" (__r)); __r; }) +#define amask(a) ({ uint64_t __r; asm ("amask %1,%0" : "=r" (__r) : "rI" (a)); __r; }) +#define implver() ({ uint64_t __r; asm ("implver %0" : "=r" (__r)); __r; }) +#define rpcc() ({ uint64_t __r; asm volatile ("rpcc %0" : "=r" (__r)); __r; }) #endif #define wh64(p) asm volatile("wh64 (%0)" : : "r"(p) : "memory") #if GNUC_PREREQ(3,3) && defined(__alpha_max__) -#define minub8 __builtin_alpha_minub8 -#define minsb8 __builtin_alpha_minsb8 -#define minuw4 __builtin_alpha_minuw4 -#define minsw4 __builtin_alpha_minsw4 -#define maxub8 __builtin_alpha_maxub8 -#define maxsb8 __builtin_alpha_maxsb8 -#define maxuw4 __builtin_alpha_maxuw4 -#define maxsw4 __builtin_alpha_maxsw4 -#define perr __builtin_alpha_perr -#define pklb __builtin_alpha_pklb -#define pkwb __builtin_alpha_pkwb -#define unpkbl __builtin_alpha_unpkbl -#define unpkbw __builtin_alpha_unpkbw +#define minub8 __builtin_alpha_minub8 +#define minsb8 __builtin_alpha_minsb8 +#define minuw4 __builtin_alpha_minuw4 +#define minsw4 __builtin_alpha_minsw4 +#define maxub8 __builtin_alpha_maxub8 +#define maxsb8 __builtin_alpha_maxsb8 +#define maxuw4 __builtin_alpha_maxuw4 +#define maxsw4 __builtin_alpha_maxsw4 +#define perr __builtin_alpha_perr +#define pklb __builtin_alpha_pklb +#define pkwb __builtin_alpha_pkwb +#define unpkbl __builtin_alpha_unpkbl +#define unpkbw __builtin_alpha_unpkbw #else #define minub8(a, b) ({ uint64_t __r; asm (".arch ev6; minub8 %r1,%2,%0" : "=r" (__r) : "%rJ" (a), "rI" (b)); __r; }) #define minsb8(a, b) ({ uint64_t __r; asm (".arch ev6; minsb8 %r1,%2,%0" : "=r" (__r) : "%rJ" (a), "rI" (b)); __r; }) @@ -143,13 +143,13 @@ #define maxuw4(a, b) ({ uint64_t __r; asm (".arch ev6; maxuw4 %r1,%2,%0" : "=r" (__r) : "%rJ" (a), "rI" (b)); __r; }) #define maxsw4(a, b) ({ uint64_t __r; asm (".arch ev6; maxsw4 %r1,%2,%0" : "=r" (__r) : "%rJ" (a), "rI" (b)); __r; }) #define perr(a, b) ({ uint64_t __r; asm (".arch ev6; perr %r1,%r2,%0" : "=r" (__r) : "%rJ" (a), "rJ" (b)); __r; }) -#define pklb(a) ({ uint64_t __r; asm (".arch ev6; pklb %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) -#define pkwb(a) ({ uint64_t __r; asm (".arch ev6; pkwb %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) -#define unpkbl(a) ({ uint64_t __r; asm (".arch ev6; unpkbl %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) -#define unpkbw(a) ({ uint64_t __r; asm (".arch ev6; unpkbw %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) +#define pklb(a) ({ uint64_t __r; asm (".arch ev6; pklb %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) +#define pkwb(a) ({ uint64_t __r; asm (".arch ev6; pkwb %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) +#define unpkbl(a) ({ uint64_t __r; asm (".arch ev6; unpkbl %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) +#define unpkbw(a) ({ uint64_t __r; asm (".arch ev6; unpkbw %r1,%0" : "=r" (__r) : "rJ" (a)); __r; }) #endif -#elif defined(__DECC) /* Digital/Compaq/hp "ccc" compiler */ +#elif defined(__DECC) /* Digital/Compaq/hp "ccc" compiler */ #include <c_asm.h> #define ldq(p) (*(const uint64_t *) (p)) @@ -157,7 +157,7 @@ #define stq(l, p) do { *(uint64_t *) (p) = (l); } while (0) #define stl(l, p) do { *(int32_t *) (p) = (l); } while (0) #define ldq_u(a) asm ("ldq_u %v0,0(%a0)", a) -#define uldq(a) (*(const __unaligned uint64_t *) (a)) +#define uldq(a) (*(const __unaligned uint64_t *) (a)) #define cmpbge(a, b) asm ("cmpbge %a0,%a1,%v0", a, b) #define extql(a, b) asm ("extql %a0,%a1,%v0", a, b) #define extwl(a, b) asm ("extwl %a0,%a1,%v0", a, b) @@ -166,7 +166,7 @@ #define zapnot(a, b) asm ("zapnot %a0,%a1,%v0", a, b) #define amask(a) asm ("amask %a0,%v0", a) #define implver() asm ("implver %v0") -#define rpcc() asm ("rpcc %v0") +#define rpcc() asm ("rpcc %v0") #define minub8(a, b) asm ("minub8 %a0,%a1,%v0", a, b) #define minsb8(a, b) asm ("minsb8 %a0,%a1,%v0", a, b) #define minuw4(a, b) asm ("minuw4 %a0,%a1,%v0", a, b)
--- a/alpha/dsputil_alpha_asm.S Wed Dec 21 17:50:40 2005 +0000 +++ b/alpha/dsputil_alpha_asm.S Thu Dec 22 01:10:11 2005 +0000 @@ -71,7 +71,7 @@ addq a1, a2, a1 nop - ldq_u t4, 0(a1) + ldq_u t4, 0(a1) ldq_u t5, 8(a1) addq a1, a2, a1 nop @@ -120,20 +120,20 @@ addq a1, a2, a1 ldq t3, 0(a1) - addq a0, a2, t4 - addq a1, a2, a1 - addq t4, a2, t5 - subq a3, 4, a3 + addq a0, a2, t4 + addq a1, a2, a1 + addq t4, a2, t5 + subq a3, 4, a3 - stq t0, 0(a0) - addq t5, a2, t6 - stq t1, 0(t4) - addq t6, a2, a0 + stq t0, 0(a0) + addq t5, a2, t6 + stq t1, 0(t4) + addq t6, a2, a0 - stq t2, 0(t5) - stq t3, 0(t6) + stq t2, 0(t5) + stq t3, 0(t6) - bne a3, $aligned + bne a3, $aligned ret .end put_pixels_axp_asm
--- a/alpha/motion_est_alpha.c Wed Dec 21 17:50:40 2005 +0000 +++ b/alpha/motion_est_alpha.c Thu Dec 22 01:10:11 2005 +0000 @@ -116,7 +116,7 @@ return result; } -#if 0 /* now done in assembly */ +#if 0 /* now done in assembly */ int pix_abs16x16_mvi(uint8_t *pix1, uint8_t *pix2, int line_size) { int result = 0;
--- a/alpha/simple_idct_alpha.c Wed Dec 21 17:50:40 2005 +0000 +++ b/alpha/simple_idct_alpha.c Thu Dec 22 01:10:11 2005 +0000 @@ -285,7 +285,7 @@ stq(v, block + 1 * 4); stq(w, block + 2 * 4); stq(w, block + 3 * 4); - block += 4 * 4; + block += 4 * 4; } } else { for (i = 0; i < 8; i++)
--- a/amr.c Wed Dec 21 17:50:40 2005 +0000 +++ b/amr.c Thu Dec 22 01:10:11 2005 +0000 @@ -301,7 +301,7 @@ static int amr_nb_encode_frame(AVCodecContext *avctx, - unsigned char *frame/*out*/, int buf_size, void *data/*in*/) + unsigned char *frame/*out*/, int buf_size, void *data/*in*/) { short serial_data[250] = {0}; @@ -440,7 +440,7 @@ } static int amr_nb_encode_frame(AVCodecContext *avctx, - unsigned char *frame/*out*/, int buf_size, void *data/*in*/) + unsigned char *frame/*out*/, int buf_size, void *data/*in*/) { AMRContext *s = (AMRContext*)avctx->priv_data; int written; @@ -584,7 +584,7 @@ } static int amr_wb_encode_frame(AVCodecContext *avctx, - unsigned char *frame/*out*/, int buf_size, void *data/*in*/) + unsigned char *frame/*out*/, int buf_size, void *data/*in*/) { AMRWBContext *s = (AMRWBContext*) avctx->priv_data; int size = E_IF_encode(s->state, s->mode, data, frame, s->allow_dtx);
--- a/armv4l/dsputil_arm.c Wed Dec 21 17:50:40 2005 +0000 +++ b/armv4l/dsputil_arm.c Thu Dec 22 01:10:11 2005 +0000 @@ -205,13 +205,13 @@ #endif c->idct_put= j_rev_dct_ARM_put; c->idct_add= j_rev_dct_ARM_add; - c->idct = j_rev_dct_ARM; + c->idct = j_rev_dct_ARM; c->idct_permutation_type= FF_LIBMPEG2_IDCT_PERM;/* FF_NO_IDCT_PERM */ } else if (idct_algo==FF_IDCT_SIMPLEARM){ - c->idct_put= simple_idct_ARM_put; - c->idct_add= simple_idct_ARM_add; - c->idct = simple_idct_ARM; - c->idct_permutation_type= FF_NO_IDCT_PERM; + c->idct_put= simple_idct_ARM_put; + c->idct_add= simple_idct_ARM_add; + c->idct = simple_idct_ARM; + c->idct_permutation_type= FF_NO_IDCT_PERM; #ifdef HAVE_IPP } else if (idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_IPP){ #else
--- a/armv4l/dsputil_iwmmxt.c Wed Dec 21 17:50:40 2005 +0000 +++ b/armv4l/dsputil_iwmmxt.c Thu Dec 22 01:10:11 2005 +0000 @@ -138,10 +138,10 @@ mm_flags = mm_support(); if (avctx->dsp_mask) { - if (avctx->dsp_mask & FF_MM_FORCE) - mm_flags |= (avctx->dsp_mask & 0xffff); - else - mm_flags &= ~(avctx->dsp_mask & 0xffff); + if (avctx->dsp_mask & FF_MM_FORCE) + mm_flags |= (avctx->dsp_mask & 0xffff); + else + mm_flags &= ~(avctx->dsp_mask & 0xffff); } if (!(mm_flags & MM_IWMMXT)) return;
--- a/armv4l/jrevdct_arm.S Wed Dec 21 17:50:40 2005 +0000 +++ b/armv4l/jrevdct_arm.S Thu Dec 22 01:10:11 2005 +0000 @@ -1,6 +1,6 @@ /* C-like prototype : - void j_rev_dct_ARM(DCTBLOCK data) + void j_rev_dct_ARM(DCTBLOCK data) With DCTBLOCK being a pointer to an array of 64 'signed shorts' @@ -51,336 +51,336 @@ #define FIX_M_1_961570560_ID 40 #define FIX_M_2_562915447_ID 44 #define FIX_0xFFFF_ID 48 - .text - .align + .text + .align - .global j_rev_dct_ARM + .global j_rev_dct_ARM j_rev_dct_ARM: - stmdb sp!, { r4 - r12, lr } @ all callee saved regs + stmdb sp!, { r4 - r12, lr } @ all callee saved regs - sub sp, sp, #4 @ reserve some space on the stack - str r0, [ sp ] @ save the DCT pointer to the stack + sub sp, sp, #4 @ reserve some space on the stack + str r0, [ sp ] @ save the DCT pointer to the stack - mov lr, r0 @ lr = pointer to the current row - mov r12, #8 @ r12 = row-counter - add r11, pc, #(const_array-.-8) @ r11 = base pointer to the constants array + mov lr, r0 @ lr = pointer to the current row + mov r12, #8 @ r12 = row-counter + add r11, pc, #(const_array-.-8) @ r11 = base pointer to the constants array row_loop: - ldrsh r0, [lr, # 0] @ r0 = 'd0' - ldrsh r1, [lr, # 8] @ r1 = 'd1' + ldrsh r0, [lr, # 0] @ r0 = 'd0' + ldrsh r1, [lr, # 8] @ r1 = 'd1' - @ Optimization for row that have all items except the first set to 0 - @ (this works as the DCTELEMS are always 4-byte aligned) - ldr r5, [lr, # 0] - ldr r2, [lr, # 4] - ldr r3, [lr, # 8] - ldr r4, [lr, #12] - orr r3, r3, r4 - orr r3, r3, r2 - orrs r5, r3, r5 - beq end_of_row_loop @ nothing to be done as ALL of them are '0' - orrs r2, r3, r1 - beq empty_row + @ Optimization for row that have all items except the first set to 0 + @ (this works as the DCTELEMS are always 4-byte aligned) + ldr r5, [lr, # 0] + ldr r2, [lr, # 4] + ldr r3, [lr, # 8] + ldr r4, [lr, #12] + orr r3, r3, r4 + orr r3, r3, r2 + orrs r5, r3, r5 + beq end_of_row_loop @ nothing to be done as ALL of them are '0' + orrs r2, r3, r1 + beq empty_row - ldrsh r2, [lr, # 2] @ r2 = 'd2' - ldrsh r4, [lr, # 4] @ r4 = 'd4' - ldrsh r6, [lr, # 6] @ r6 = 'd6' + ldrsh r2, [lr, # 2] @ r2 = 'd2' + ldrsh r4, [lr, # 4] @ r4 = 'd4' + ldrsh r6, [lr, # 6] @ r6 = 'd6' - ldr r3, [r11, #FIX_0_541196100_ID] - add r7, r2, r6 - ldr r5, [r11, #FIX_M_1_847759065_ID] - mul r7, r3, r7 @ r7 = z1 - ldr r3, [r11, #FIX_0_765366865_ID] - mla r6, r5, r6, r7 @ r6 = tmp2 - add r5, r0, r4 @ r5 = tmp0 - mla r2, r3, r2, r7 @ r2 = tmp3 - sub r3, r0, r4 @ r3 = tmp1 + ldr r3, [r11, #FIX_0_541196100_ID] + add r7, r2, r6 + ldr r5, [r11, #FIX_M_1_847759065_ID] + mul r7, r3, r7 @ r7 = z1 + ldr r3, [r11, #FIX_0_765366865_ID] + mla r6, r5, r6, r7 @ r6 = tmp2 + add r5, r0, r4 @ r5 = tmp0 + mla r2, r3, r2, r7 @ r2 = tmp3 + sub r3, r0, r4 @ r3 = tmp1 - add r0, r2, r5, lsl #13 @ r0 = tmp10 - rsb r2, r2, r5, lsl #13 @ r2 = tmp13 - add r4, r6, r3, lsl #13 @ r4 = tmp11 - rsb r3, r6, r3, lsl #13 @ r3 = tmp12 + add r0, r2, r5, lsl #13 @ r0 = tmp10 + rsb r2, r2, r5, lsl #13 @ r2 = tmp13 + add r4, r6, r3, lsl #13 @ r4 = tmp11 + rsb r3, r6, r3, lsl #13 @ r3 = tmp12 - stmdb sp!, { r0, r2, r3, r4 } @ save on the stack tmp10, tmp13, tmp12, tmp11 + stmdb sp!, { r0, r2, r3, r4 } @ save on the stack tmp10, tmp13, tmp12, tmp11 - ldrsh r3, [lr, #10] @ r3 = 'd3' - ldrsh r5, [lr, #12] @ r5 = 'd5' - ldrsh r7, [lr, #14] @ r7 = 'd7' + ldrsh r3, [lr, #10] @ r3 = 'd3' + ldrsh r5, [lr, #12] @ r5 = 'd5' + ldrsh r7, [lr, #14] @ r7 = 'd7' - add r0, r3, r5 @ r0 = 'z2' - add r2, r1, r7 @ r2 = 'z1' - add r4, r3, r7 @ r4 = 'z3' - add r6, r1, r5 @ r6 = 'z4' - ldr r9, [r11, #FIX_1_175875602_ID] - add r8, r4, r6 @ r8 = z3 + z4 - ldr r10, [r11, #FIX_M_0_899976223_ID] - mul r8, r9, r8 @ r8 = 'z5' - ldr r9, [r11, #FIX_M_2_562915447_ID] - mul r2, r10, r2 @ r2 = 'z1' - ldr r10, [r11, #FIX_M_1_961570560_ID] - mul r0, r9, r0 @ r0 = 'z2' - ldr r9, [r11, #FIX_M_0_390180644_ID] - mla r4, r10, r4, r8 @ r4 = 'z3' - ldr r10, [r11, #FIX_0_298631336_ID] - mla r6, r9, r6, r8 @ r6 = 'z4' - ldr r9, [r11, #FIX_2_053119869_ID] - mla r7, r10, r7, r2 @ r7 = tmp0 + z1 - ldr r10, [r11, #FIX_3_072711026_ID] - mla r5, r9, r5, r0 @ r5 = tmp1 + z2 - ldr r9, [r11, #FIX_1_501321110_ID] - mla r3, r10, r3, r0 @ r3 = tmp2 + z2 - add r7, r7, r4 @ r7 = tmp0 - mla r1, r9, r1, r2 @ r1 = tmp3 + z1 - add r5, r5, r6 @ r5 = tmp1 - add r3, r3, r4 @ r3 = tmp2 - add r1, r1, r6 @ r1 = tmp3 + add r0, r3, r5 @ r0 = 'z2' + add r2, r1, r7 @ r2 = 'z1' + add r4, r3, r7 @ r4 = 'z3' + add r6, r1, r5 @ r6 = 'z4' + ldr r9, [r11, #FIX_1_175875602_ID] + add r8, r4, r6 @ r8 = z3 + z4 + ldr r10, [r11, #FIX_M_0_899976223_ID] + mul r8, r9, r8 @ r8 = 'z5' + ldr r9, [r11, #FIX_M_2_562915447_ID] + mul r2, r10, r2 @ r2 = 'z1' + ldr r10, [r11, #FIX_M_1_961570560_ID] + mul r0, r9, r0 @ r0 = 'z2' + ldr r9, [r11, #FIX_M_0_390180644_ID] + mla r4, r10, r4, r8 @ r4 = 'z3' + ldr r10, [r11, #FIX_0_298631336_ID] + mla r6, r9, r6, r8 @ r6 = 'z4' + ldr r9, [r11, #FIX_2_053119869_ID] + mla r7, r10, r7, r2 @ r7 = tmp0 + z1 + ldr r10, [r11, #FIX_3_072711026_ID] + mla r5, r9, r5, r0 @ r5 = tmp1 + z2 + ldr r9, [r11, #FIX_1_501321110_ID] + mla r3, r10, r3, r0 @ r3 = tmp2 + z2 + add r7, r7, r4 @ r7 = tmp0 + mla r1, r9, r1, r2 @ r1 = tmp3 + z1 + add r5, r5, r6 @ r5 = tmp1 + add r3, r3, r4 @ r3 = tmp2 + add r1, r1, r6 @ r1 = tmp3 - ldmia sp!, { r0, r2, r4, r6 } @ r0 = tmp10 / r2 = tmp13 / r4 = tmp12 / r6 = tmp11 - @ r1 = tmp3 / r3 = tmp2 / r5 = tmp1 / r7 = tmp0 + ldmia sp!, { r0, r2, r4, r6 } @ r0 = tmp10 / r2 = tmp13 / r4 = tmp12 / r6 = tmp11 + @ r1 = tmp3 / r3 = tmp2 / r5 = tmp1 / r7 = tmp0 - @ Compute DESCALE(tmp10 + tmp3, CONST_BITS-PASS1_BITS) - add r8, r0, r1 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, # 0] + @ Compute DESCALE(tmp10 + tmp3, CONST_BITS-PASS1_BITS) + add r8, r0, r1 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, # 0] - @ Compute DESCALE(tmp10 - tmp3, CONST_BITS-PASS1_BITS) - sub r8, r0, r1 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, #14] + @ Compute DESCALE(tmp10 - tmp3, CONST_BITS-PASS1_BITS) + sub r8, r0, r1 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, #14] - @ Compute DESCALE(tmp11 + tmp2, CONST_BITS-PASS1_BITS) - add r8, r6, r3 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, # 2] + @ Compute DESCALE(tmp11 + tmp2, CONST_BITS-PASS1_BITS) + add r8, r6, r3 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, # 2] - @ Compute DESCALE(tmp11 - tmp2, CONST_BITS-PASS1_BITS) - sub r8, r6, r3 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, #12] + @ Compute DESCALE(tmp11 - tmp2, CONST_BITS-PASS1_BITS) + sub r8, r6, r3 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, #12] - @ Compute DESCALE(tmp12 + tmp1, CONST_BITS-PASS1_BITS) - add r8, r4, r5 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, # 4] + @ Compute DESCALE(tmp12 + tmp1, CONST_BITS-PASS1_BITS) + add r8, r4, r5 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, # 4] - @ Compute DESCALE(tmp12 - tmp1, CONST_BITS-PASS1_BITS) - sub r8, r4, r5 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, #10] + @ Compute DESCALE(tmp12 - tmp1, CONST_BITS-PASS1_BITS) + sub r8, r4, r5 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, #10] - @ Compute DESCALE(tmp13 + tmp0, CONST_BITS-PASS1_BITS) - add r8, r2, r7 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, # 6] + @ Compute DESCALE(tmp13 + tmp0, CONST_BITS-PASS1_BITS) + add r8, r2, r7 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, # 6] - @ Compute DESCALE(tmp13 - tmp0, CONST_BITS-PASS1_BITS) - sub r8, r2, r7 - add r8, r8, #(1<<10) - mov r8, r8, asr #11 - strh r8, [lr, # 8] + @ Compute DESCALE(tmp13 - tmp0, CONST_BITS-PASS1_BITS) + sub r8, r2, r7 + add r8, r8, #(1<<10) + mov r8, r8, asr #11 + strh r8, [lr, # 8] - @ End of row loop - add lr, lr, #16 - subs r12, r12, #1 - bne row_loop - beq start_column_loop + @ End of row loop + add lr, lr, #16 + subs r12, r12, #1 + bne row_loop + beq start_column_loop empty_row: - ldr r1, [r11, #FIX_0xFFFF_ID] - mov r0, r0, lsl #2 - and r0, r0, r1 - add r0, r0, r0, lsl #16 - str r0, [lr, # 0] - str r0, [lr, # 4] - str r0, [lr, # 8] - str r0, [lr, #12] + ldr r1, [r11, #FIX_0xFFFF_ID] + mov r0, r0, lsl #2 + and r0, r0, r1 + add r0, r0, r0, lsl #16 + str r0, [lr, # 0] + str r0, [lr, # 4] + str r0, [lr, # 8] + str r0, [lr, #12] end_of_row_loop: - @ End of loop - add lr, lr, #16 - subs r12, r12, #1 - bne row_loop + @ End of loop + add lr, lr, #16 + subs r12, r12, #1 + bne row_loop start_column_loop: - @ Start of column loop - ldr lr, [ sp ] - mov r12, #8 + @ Start of column loop + ldr lr, [ sp ] + mov r12, #8 column_loop: - ldrsh r0, [lr, #( 0*8)] @ r0 = 'd0' - ldrsh r2, [lr, #( 4*8)] @ r2 = 'd2' - ldrsh r4, [lr, #( 8*8)] @ r4 = 'd4' - ldrsh r6, [lr, #(12*8)] @ r6 = 'd6' + ldrsh r0, [lr, #( 0*8)] @ r0 = 'd0' + ldrsh r2, [lr, #( 4*8)] @ r2 = 'd2' + ldrsh r4, [lr, #( 8*8)] @ r4 = 'd4' + ldrsh r6, [lr, #(12*8)] @ r6 = 'd6' - ldr r3, [r11, #FIX_0_541196100_ID] - add r1, r2, r6 - ldr r5, [r11, #FIX_M_1_847759065_ID] - mul r1, r3, r1 @ r1 = z1 - ldr r3, [r11, #FIX_0_765366865_ID] - mla r6, r5, r6, r1 @ r6 = tmp2 - add r5, r0, r4 @ r5 = tmp0 - mla r2, r3, r2, r1 @ r2 = tmp3 - sub r3, r0, r4 @ r3 = tmp1 + ldr r3, [r11, #FIX_0_541196100_ID] + add r1, r2, r6 + ldr r5, [r11, #FIX_M_1_847759065_ID] + mul r1, r3, r1 @ r1 = z1 + ldr r3, [r11, #FIX_0_765366865_ID] + mla r6, r5, r6, r1 @ r6 = tmp2 + add r5, r0, r4 @ r5 = tmp0 + mla r2, r3, r2, r1 @ r2 = tmp3 + sub r3, r0, r4 @ r3 = tmp1 - add r0, r2, r5, lsl #13 @ r0 = tmp10 - rsb r2, r2, r5, lsl #13 @ r2 = tmp13 - add r4, r6, r3, lsl #13 @ r4 = tmp11 - rsb r6, r6, r3, lsl #13 @ r6 = tmp12 + add r0, r2, r5, lsl #13 @ r0 = tmp10 + rsb r2, r2, r5, lsl #13 @ r2 = tmp13 + add r4, r6, r3, lsl #13 @ r4 = tmp11 + rsb r6, r6, r3, lsl #13 @ r6 = tmp12 - ldrsh r1, [lr, #( 2*8)] @ r1 = 'd1' - ldrsh r3, [lr, #( 6*8)] @ r3 = 'd3' - ldrsh r5, [lr, #(10*8)] @ r5 = 'd5' - ldrsh r7, [lr, #(14*8)] @ r7 = 'd7' + ldrsh r1, [lr, #( 2*8)] @ r1 = 'd1' + ldrsh r3, [lr, #( 6*8)] @ r3 = 'd3' + ldrsh r5, [lr, #(10*8)] @ r5 = 'd5' + ldrsh r7, [lr, #(14*8)] @ r7 = 'd7' - @ Check for empty odd column (happens about 20 to 25 % of the time according to my stats) - orr r9, r1, r3 - orr r10, r5, r7 - orrs r10, r9, r10 - beq empty_odd_column + @ Check for empty odd column (happens about 20 to 25 % of the time according to my stats) + orr r9, r1, r3 + orr r10, r5, r7 + orrs r10, r9, r10 + beq empty_odd_column - stmdb sp!, { r0, r2, r4, r6 } @ save on the stack tmp10, tmp13, tmp12, tmp11 + stmdb sp!, { r0, r2, r4, r6 } @ save on the stack tmp10, tmp13, tmp12, tmp11 - add r0, r3, r5 @ r0 = 'z2' - add r2, r1, r7 @ r2 = 'z1' - add r4, r3, r7 @ r4 = 'z3' - add r6, r1, r5 @ r6 = 'z4' - ldr r9, [r11, #FIX_1_175875602_ID] - add r8, r4, r6 - ldr r10, [r11, #FIX_M_0_899976223_ID] - mul r8, r9, r8 @ r8 = 'z5' - ldr r9, [r11, #FIX_M_2_562915447_ID] - mul r2, r10, r2 @ r2 = 'z1' - ldr r10, [r11, #FIX_M_1_961570560_ID] - mul r0, r9, r0 @ r0 = 'z2' - ldr r9, [r11, #FIX_M_0_390180644_ID] - mla r4, r10, r4, r8 @ r4 = 'z3' - ldr r10, [r11, #FIX_0_298631336_ID] - mla r6, r9, r6, r8 @ r6 = 'z4' - ldr r9, [r11, #FIX_2_053119869_ID] - mla r7, r10, r7, r2 @ r7 = tmp0 + z1 - ldr r10, [r11, #FIX_3_072711026_ID] - mla r5, r9, r5, r0 @ r5 = tmp1 + z2 - ldr r9, [r11, #FIX_1_501321110_ID] - mla r3, r10, r3, r0 @ r3 = tmp2 + z2 - add r7, r7, r4 @ r7 = tmp0 - mla r1, r9, r1, r2 @ r1 = tmp3 + z1 - add r5, r5, r6 @ r5 = tmp1 - add r3, r3, r4 @ r3 = tmp2 - add r1, r1, r6 @ r1 = tmp3 + add r0, r3, r5 @ r0 = 'z2' + add r2, r1, r7 @ r2 = 'z1' + add r4, r3, r7 @ r4 = 'z3' + add r6, r1, r5 @ r6 = 'z4' + ldr r9, [r11, #FIX_1_175875602_ID] + add r8, r4, r6 + ldr r10, [r11, #FIX_M_0_899976223_ID] + mul r8, r9, r8 @ r8 = 'z5' + ldr r9, [r11, #FIX_M_2_562915447_ID] + mul r2, r10, r2 @ r2 = 'z1' + ldr r10, [r11, #FIX_M_1_961570560_ID] + mul r0, r9, r0 @ r0 = 'z2' + ldr r9, [r11, #FIX_M_0_390180644_ID] + mla r4, r10, r4, r8 @ r4 = 'z3' + ldr r10, [r11, #FIX_0_298631336_ID] + mla r6, r9, r6, r8 @ r6 = 'z4' + ldr r9, [r11, #FIX_2_053119869_ID] + mla r7, r10, r7, r2 @ r7 = tmp0 + z1 + ldr r10, [r11, #FIX_3_072711026_ID] + mla r5, r9, r5, r0 @ r5 = tmp1 + z2 + ldr r9, [r11, #FIX_1_501321110_ID] + mla r3, r10, r3, r0 @ r3 = tmp2 + z2 + add r7, r7, r4 @ r7 = tmp0 + mla r1, r9, r1, r2 @ r1 = tmp3 + z1 + add r5, r5, r6 @ r5 = tmp1 + add r3, r3, r4 @ r3 = tmp2 + add r1, r1, r6 @ r1 = tmp3 - ldmia sp!, { r0, r2, r4, r6 } @ r0 = tmp10 / r2 = tmp13 / r4 = tmp11 / r6 = tmp12 - @ r1 = tmp3 / r3 = tmp2 / r5 = tmp1 / r7 = tmp0 + ldmia sp!, { r0, r2, r4, r6 } @ r0 = tmp10 / r2 = tmp13 / r4 = tmp11 / r6 = tmp12 + @ r1 = tmp3 / r3 = tmp2 / r5 = tmp1 / r7 = tmp0 - @ Compute DESCALE(tmp10 + tmp3, CONST_BITS+PASS1_BITS+3) - add r8, r0, r1 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #( 0*8)] + @ Compute DESCALE(tmp10 + tmp3, CONST_BITS+PASS1_BITS+3) + add r8, r0, r1 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #( 0*8)] - @ Compute DESCALE(tmp10 - tmp3, CONST_BITS+PASS1_BITS+3) - sub r8, r0, r1 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #(14*8)] + @ Compute DESCALE(tmp10 - tmp3, CONST_BITS+PASS1_BITS+3) + sub r8, r0, r1 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #(14*8)] - @ Compute DESCALE(tmp11 + tmp2, CONST_BITS+PASS1_BITS+3) - add r8, r4, r3 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #( 2*8)] + @ Compute DESCALE(tmp11 + tmp2, CONST_BITS+PASS1_BITS+3) + add r8, r4, r3 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #( 2*8)] - @ Compute DESCALE(tmp11 - tmp2, CONST_BITS+PASS1_BITS+3) - sub r8, r4, r3 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #(12*8)] + @ Compute DESCALE(tmp11 - tmp2, CONST_BITS+PASS1_BITS+3) + sub r8, r4, r3 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #(12*8)] - @ Compute DESCALE(tmp12 + tmp1, CONST_BITS+PASS1_BITS+3) - add r8, r6, r5 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #( 4*8)] + @ Compute DESCALE(tmp12 + tmp1, CONST_BITS+PASS1_BITS+3) + add r8, r6, r5 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #( 4*8)] - @ Compute DESCALE(tmp12 - tmp1, CONST_BITS+PASS1_BITS+3) - sub r8, r6, r5 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #(10*8)] + @ Compute DESCALE(tmp12 - tmp1, CONST_BITS+PASS1_BITS+3) + sub r8, r6, r5 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #(10*8)] - @ Compute DESCALE(tmp13 + tmp0, CONST_BITS+PASS1_BITS+3) - add r8, r2, r7 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #( 6*8)] + @ Compute DESCALE(tmp13 + tmp0, CONST_BITS+PASS1_BITS+3) + add r8, r2, r7 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #( 6*8)] - @ Compute DESCALE(tmp13 - tmp0, CONST_BITS+PASS1_BITS+3) - sub r8, r2, r7 - add r8, r8, #(1<<17) - mov r8, r8, asr #18 - strh r8, [lr, #( 8*8)] + @ Compute DESCALE(tmp13 - tmp0, CONST_BITS+PASS1_BITS+3) + sub r8, r2, r7 + add r8, r8, #(1<<17) + mov r8, r8, asr #18 + strh r8, [lr, #( 8*8)] - @ End of row loop - add lr, lr, #2 - subs r12, r12, #1 - bne column_loop - beq the_end + @ End of row loop + add lr, lr, #2 + subs r12, r12, #1 + bne column_loop + beq the_end empty_odd_column: - @ Compute DESCALE(tmp10 + tmp3, CONST_BITS+PASS1_BITS+3) - @ Compute DESCALE(tmp10 - tmp3, CONST_BITS+PASS1_BITS+3) - add r0, r0, #(1<<17) - mov r0, r0, asr #18 - strh r0, [lr, #( 0*8)] - strh r0, [lr, #(14*8)] + @ Compute DESCALE(tmp10 + tmp3, CONST_BITS+PASS1_BITS+3) + @ Compute DESCALE(tmp10 - tmp3, CONST_BITS+PASS1_BITS+3) + add r0, r0, #(1<<17) + mov r0, r0, asr #18 + strh r0, [lr, #( 0*8)] + strh r0, [lr, #(14*8)] - @ Compute DESCALE(tmp11 + tmp2, CONST_BITS+PASS1_BITS+3) - @ Compute DESCALE(tmp11 - tmp2, CONST_BITS+PASS1_BITS+3) - add r4, r4, #(1<<17) - mov r4, r4, asr #18 - strh r4, [lr, #( 2*8)] - strh r4, [lr, #(12*8)] + @ Compute DESCALE(tmp11 + tmp2, CONST_BITS+PASS1_BITS+3) + @ Compute DESCALE(tmp11 - tmp2, CONST_BITS+PASS1_BITS+3) + add r4, r4, #(1<<17) + mov r4, r4, asr #18 + strh r4, [lr, #( 2*8)] + strh r4, [lr, #(12*8)] - @ Compute DESCALE(tmp12 + tmp1, CONST_BITS+PASS1_BITS+3) - @ Compute DESCALE(tmp12 - tmp1, CONST_BITS+PASS1_BITS+3) - add r6, r6, #(1<<17) - mov r6, r6, asr #18 - strh r6, [lr, #( 4*8)] - strh r6, [lr, #(10*8)] + @ Compute DESCALE(tmp12 + tmp1, CONST_BITS+PASS1_BITS+3) + @ Compute DESCALE(tmp12 - tmp1, CONST_BITS+PASS1_BITS+3) + add r6, r6, #(1<<17) + mov r6, r6, asr #18 + strh r6, [lr, #( 4*8)] + strh r6, [lr, #(10*8)] - @ Compute DESCALE(tmp13 + tmp0, CONST_BITS+PASS1_BITS+3) - @ Compute DESCALE(tmp13 - tmp0, CONST_BITS+PASS1_BITS+3) - add r2, r2, #(1<<17) - mov r2, r2, asr #18 - strh r2, [lr, #( 6*8)] - strh r2, [lr, #( 8*8)] + @ Compute DESCALE(tmp13 + tmp0, CONST_BITS+PASS1_BITS+3) + @ Compute DESCALE(tmp13 - tmp0, CONST_BITS+PASS1_BITS+3) + add r2, r2, #(1<<17) + mov r2, r2, asr #18 + strh r2, [lr, #( 6*8)] + strh r2, [lr, #( 8*8)] - @ End of row loop - add lr, lr, #2 - subs r12, r12, #1 - bne column_loop + @ End of row loop + add lr, lr, #2 + subs r12, r12, #1 + bne column_loop the_end: - @ The end.... - add sp, sp, #4 - ldmia sp!, { r4 - r12, pc } @ restore callee saved regs and return + @ The end.... + add sp, sp, #4 + ldmia sp!, { r4 - r12, pc } @ restore callee saved regs and return const_array: - .align - .word FIX_0_298631336 - .word FIX_0_541196100 - .word FIX_0_765366865 - .word FIX_1_175875602 - .word FIX_1_501321110 - .word FIX_2_053119869 - .word FIX_3_072711026 - .word FIX_M_0_390180644 - .word FIX_M_0_899976223 - .word FIX_M_1_847759065 - .word FIX_M_1_961570560 - .word FIX_M_2_562915447 - .word FIX_0xFFFF + .align + .word FIX_0_298631336 + .word FIX_0_541196100 + .word FIX_0_765366865 + .word FIX_1_175875602 + .word FIX_1_501321110 + .word FIX_2_053119869 + .word FIX_3_072711026 + .word FIX_M_0_390180644 + .word FIX_M_0_899976223 + .word FIX_M_1_847759065 + .word FIX_M_1_961570560 + .word FIX_M_2_562915447 + .word FIX_0xFFFF
--- a/armv4l/simple_idct_arm.S Wed Dec 21 17:50:40 2005 +0000 +++ b/armv4l/simple_idct_arm.S Thu Dec 22 01:10:11 2005 +0000 @@ -51,9 +51,9 @@ #define COL_SHIFTED_1 524288 /* 1<< (COL_SHIFT-1) */ - .text - .align - .global simple_idct_ARM + .text + .align + .global simple_idct_ARM simple_idct_ARM: @@ void simple_idct_ARM(int16_t *block) @@ -120,8 +120,8 @@ ldr r11, [r12, #offW7] @ R11=W7 mul r5, r10, r7 @ R5=W5*ROWr16[1]=b2 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) mul r7, r11, r7 @ R7=W7*ROWr16[1]=b3 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) - teq r2, #0 @ if null avoid muls - mlane r0, r9, r2, r0 @ R0+=W3*ROWr16[3]=b0 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) + teq r2, #0 @ if null avoid muls + mlane r0, r9, r2, r0 @ R0+=W3*ROWr16[3]=b0 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) rsbne r2, r2, #0 @ R2=-ROWr16[3] mlane r1, r11, r2, r1 @ R1-=W7*ROWr16[3]=b1 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) mlane r5, r8, r2, r5 @ R5-=W1*ROWr16[3]=b2 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) @@ -147,7 +147,7 @@ @@ MAC16(b3, -W1, row[7]); @@ MAC16(b1, -W5, row[7]); mov r3, r3, asr #16 @ R3=ROWr16[5] - teq r3, #0 @ if null avoid muls + teq r3, #0 @ if null avoid muls mlane r0, r10, r3, r0 @ R0+=W5*ROWr16[5]=b0 mov r4, r4, asr #16 @ R4=ROWr16[7] mlane r5, r11, r3, r5 @ R5+=W7*ROWr16[5]=b2 @@ -155,7 +155,7 @@ rsbne r3, r3, #0 @ R3=-ROWr16[5] mlane r1, r8, r3, r1 @ R7-=W1*ROWr16[5]=b1 @@ R3 is free now - teq r4, #0 @ if null avoid muls + teq r4, #0 @ if null avoid muls mlane r0, r11, r4, r0 @ R0+=W7*ROWr16[7]=b0 mlane r5, r9, r4, r5 @ R5+=W3*ROWr16[7]=b2 rsbne r4, r4, #0 @ R4=-ROWr16[7] @@ -187,7 +187,7 @@ teq r2, #0 beq __end_bef_a_evaluation - add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) + add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) mul r11, r8, r4 @ R11=W2*ROWr16[2] sub r4, r6, r11 @ R4=a0-W2*ROWr16[2] (a3) add r6, r6, r11 @ R6=a0+W2*ROWr16[2] (a0) @@ -203,7 +203,7 @@ @@ a2 -= W4*row[4] @@ a3 += W4*row[4] ldrsh r11, [r14, #8] @ R11=ROWr16[4] - teq r11, #0 @ if null avoid muls + teq r11, #0 @ if null avoid muls mulne r11, r9, r11 @ R11=W4*ROWr16[4] @@ R9 is free now ldrsh r9, [r14, #12] @ R9=ROWr16[6] @@ -212,7 +212,7 @@ subne r3, r3, r11 @ R3-=W4*ROWr16[4] (a2) addne r4, r4, r11 @ R4+=W4*ROWr16[4] (a3) @@ W6 alone is no more useful, save W2*ROWr16[6] in it instead - teq r9, #0 @ if null avoid muls + teq r9, #0 @ if null avoid muls mulne r11, r10, r9 @ R11=W6*ROWr16[6] addne r6, r6, r11 @ R6+=W6*ROWr16[6] (a0) mulne r10, r8, r9 @ R10=W2*ROWr16[6] @@ -294,165 +294,165 @@ - @@ at this point, R0=block, R1-R11 (free) - @@ R12=__const_ptr_, R14=&block[n] - add r14, r0, #14 @ R14=&block[7], better start from the last col, and decrease the value until col=0, i.e. R14=block. + @@ at this point, R0=block, R1-R11 (free) + @@ R12=__const_ptr_, R14=&block[n] + add r14, r0, #14 @ R14=&block[7], better start from the last col, and decrease the value until col=0, i.e. R14=block. __col_loop: __b_evaluation2: - @@ at this point, R0=block (temp), R1-R11 (free) - @@ R12=__const_ptr_, R14=&block[n] - @@ proceed with b0-b3 first, followed by a0-a3 - @@ MUL16(b0, W1, col[8x1]); - @@ MUL16(b1, W3, col[8x1]); - @@ MUL16(b2, W5, col[8x1]); - @@ MUL16(b3, W7, col[8x1]); - @@ MAC16(b0, W3, col[8x3]); - @@ MAC16(b1, -W7, col[8x3]); - @@ MAC16(b2, -W1, col[8x3]); - @@ MAC16(b3, -W5, col[8x3]); - ldr r8, [r12, #offW1] @ R8=W1 - ldrsh r7, [r14, #16] - mul r0, r8, r7 @ R0=W1*ROWr16[1]=b0 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) - ldr r9, [r12, #offW3] @ R9=W3 - ldr r10, [r12, #offW5] @ R10=W5 - mul r1, r9, r7 @ R1=W3*ROWr16[1]=b1 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) - ldr r11, [r12, #offW7] @ R11=W7 - mul r5, r10, r7 @ R5=W5*ROWr16[1]=b2 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) - ldrsh r2, [r14, #48] - mul r7, r11, r7 @ R7=W7*ROWr16[1]=b3 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) - teq r2, #0 @ if 0, then avoid muls - mlane r0, r9, r2, r0 @ R0+=W3*ROWr16[3]=b0 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) - rsbne r2, r2, #0 @ R2=-ROWr16[3] - mlane r1, r11, r2, r1 @ R1-=W7*ROWr16[3]=b1 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) - mlane r5, r8, r2, r5 @ R5-=W1*ROWr16[3]=b2 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) - mlane r7, r10, r2, r7 @ R7-=W5*ROWr16[3]=b3 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) + @@ at this point, R0=block (temp), R1-R11 (free) + @@ R12=__const_ptr_, R14=&block[n] + @@ proceed with b0-b3 first, followed by a0-a3 + @@ MUL16(b0, W1, col[8x1]); + @@ MUL16(b1, W3, col[8x1]); + @@ MUL16(b2, W5, col[8x1]); + @@ MUL16(b3, W7, col[8x1]); + @@ MAC16(b0, W3, col[8x3]); + @@ MAC16(b1, -W7, col[8x3]); + @@ MAC16(b2, -W1, col[8x3]); + @@ MAC16(b3, -W5, col[8x3]); + ldr r8, [r12, #offW1] @ R8=W1 + ldrsh r7, [r14, #16] + mul r0, r8, r7 @ R0=W1*ROWr16[1]=b0 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) + ldr r9, [r12, #offW3] @ R9=W3 + ldr r10, [r12, #offW5] @ R10=W5 + mul r1, r9, r7 @ R1=W3*ROWr16[1]=b1 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) + ldr r11, [r12, #offW7] @ R11=W7 + mul r5, r10, r7 @ R5=W5*ROWr16[1]=b2 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) + ldrsh r2, [r14, #48] + mul r7, r11, r7 @ R7=W7*ROWr16[1]=b3 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle) + teq r2, #0 @ if 0, then avoid muls + mlane r0, r9, r2, r0 @ R0+=W3*ROWr16[3]=b0 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) + rsbne r2, r2, #0 @ R2=-ROWr16[3] + mlane r1, r11, r2, r1 @ R1-=W7*ROWr16[3]=b1 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) + mlane r5, r8, r2, r5 @ R5-=W1*ROWr16[3]=b2 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) + mlane r7, r10, r2, r7 @ R7-=W5*ROWr16[3]=b3 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle) - @@ at this point, R0=b0, R1=b1, R2 (free), R3 (free), R4 (free), - @@ R5=b2, R6 (free), R7=b3, R8=W1, R9=W3, R10=W5, R11=W7, - @@ R12=__const_ptr_, R14=&block[n] - @@ MAC16(b0, W5, col[5x8]); - @@ MAC16(b2, W7, col[5x8]); - @@ MAC16(b3, W3, col[5x8]); - @@ MAC16(b1, -W1, col[5x8]); - @@ MAC16(b0, W7, col[7x8]); - @@ MAC16(b2, W3, col[7x8]); - @@ MAC16(b3, -W1, col[7x8]); - @@ MAC16(b1, -W5, col[7x8]); - ldrsh r3, [r14, #80] @ R3=COLr16[5x8] - teq r3, #0 @ if 0 then avoid muls - mlane r0, r10, r3, r0 @ R0+=W5*ROWr16[5x8]=b0 - mlane r5, r11, r3, r5 @ R5+=W7*ROWr16[5x8]=b2 - mlane r7, r9, r3, r7 @ R7+=W3*ROWr16[5x8]=b3 - rsbne r3, r3, #0 @ R3=-ROWr16[5x8] - ldrsh r4, [r14, #112] @ R4=COLr16[7x8] - mlane r1, r8, r3, r1 @ R7-=W1*ROWr16[5x8]=b1 - @@ R3 is free now - teq r4, #0 @ if 0 then avoid muls - mlane r0, r11, r4, r0 @ R0+=W7*ROWr16[7x8]=b0 - mlane r5, r9, r4, r5 @ R5+=W3*ROWr16[7x8]=b2 - rsbne r4, r4, #0 @ R4=-ROWr16[7x8] - mlane r7, r8, r4, r7 @ R7-=W1*ROWr16[7x8]=b3 - mlane r1, r10, r4, r1 @ R1-=W5*ROWr16[7x8]=b1 - @@ R4 is free now + @@ at this point, R0=b0, R1=b1, R2 (free), R3 (free), R4 (free), + @@ R5=b2, R6 (free), R7=b3, R8=W1, R9=W3, R10=W5, R11=W7, + @@ R12=__const_ptr_, R14=&block[n] + @@ MAC16(b0, W5, col[5x8]); + @@ MAC16(b2, W7, col[5x8]); + @@ MAC16(b3, W3, col[5x8]); + @@ MAC16(b1, -W1, col[5x8]); + @@ MAC16(b0, W7, col[7x8]); + @@ MAC16(b2, W3, col[7x8]); + @@ MAC16(b3, -W1, col[7x8]); + @@ MAC16(b1, -W5, col[7x8]); + ldrsh r3, [r14, #80] @ R3=COLr16[5x8] + teq r3, #0 @ if 0 then avoid muls + mlane r0, r10, r3, r0 @ R0+=W5*ROWr16[5x8]=b0 + mlane r5, r11, r3, r5 @ R5+=W7*ROWr16[5x8]=b2 + mlane r7, r9, r3, r7 @ R7+=W3*ROWr16[5x8]=b3 + rsbne r3, r3, #0 @ R3=-ROWr16[5x8] + ldrsh r4, [r14, #112] @ R4=COLr16[7x8] + mlane r1, r8, r3, r1 @ R7-=W1*ROWr16[5x8]=b1 + @@ R3 is free now + teq r4, #0 @ if 0 then avoid muls + mlane r0, r11, r4, r0 @ R0+=W7*ROWr16[7x8]=b0 + mlane r5, r9, r4, r5 @ R5+=W3*ROWr16[7x8]=b2 + rsbne r4, r4, #0 @ R4=-ROWr16[7x8] + mlane r7, r8, r4, r7 @ R7-=W1*ROWr16[7x8]=b3 + mlane r1, r10, r4, r1 @ R1-=W5*ROWr16[7x8]=b1 + @@ R4 is free now __end_b_evaluation2: - @@ at this point, R0=b0, R1=b1, R2 (free), R3 (free), R4 (free), - @@ R5=b2, R6 (free), R7=b3, R8 (free), R9 (free), R10 (free), R11 (free), - @@ R12=__const_ptr_, R14=&block[n] + @@ at this point, R0=b0, R1=b1, R2 (free), R3 (free), R4 (free), + @@ R5=b2, R6 (free), R7=b3, R8 (free), R9 (free), R10 (free), R11 (free), + @@ R12=__const_ptr_, R14=&block[n] __a_evaluation2: - @@ a0 = (W4 * col[8x0]) + (1 << (COL_SHIFT - 1)); - @@ a1 = a0 + W6 * row[2]; - @@ a2 = a0 - W6 * row[2]; - @@ a3 = a0 - W2 * row[2]; - @@ a0 = a0 + W2 * row[2]; - ldrsh r6, [r14, #0] - ldr r9, [r12, #offW4] @ R9=W4 - mul r6, r9, r6 @ R6=W4*ROWr16[0] - ldr r10, [r12, #offW6] @ R10=W6 - ldrsh r4, [r14, #32] @ R4=ROWr16[2] (a3 not defined yet) - add r6, r6, #COL_SHIFTED_1 @ R6=W4*ROWr16[0] + 1<<(COL_SHIFT-1) (a0) - mul r11, r10, r4 @ R11=W6*ROWr16[2] - ldr r8, [r12, #offW2] @ R8=W2 - add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) - sub r3, r6, r11 @ R3=a0-W6*ROWr16[2] (a2) - mul r11, r8, r4 @ R11=W2*ROWr16[2] - sub r4, r6, r11 @ R4=a0-W2*ROWr16[2] (a3) - add r6, r6, r11 @ R6=a0+W2*ROWr16[2] (a0) + @@ a0 = (W4 * col[8x0]) + (1 << (COL_SHIFT - 1)); + @@ a1 = a0 + W6 * row[2]; + @@ a2 = a0 - W6 * row[2]; + @@ a3 = a0 - W2 * row[2]; + @@ a0 = a0 + W2 * row[2]; + ldrsh r6, [r14, #0] + ldr r9, [r12, #offW4] @ R9=W4 + mul r6, r9, r6 @ R6=W4*ROWr16[0] + ldr r10, [r12, #offW6] @ R10=W6 + ldrsh r4, [r14, #32] @ R4=ROWr16[2] (a3 not defined yet) + add r6, r6, #COL_SHIFTED_1 @ R6=W4*ROWr16[0] + 1<<(COL_SHIFT-1) (a0) + mul r11, r10, r4 @ R11=W6*ROWr16[2] + ldr r8, [r12, #offW2] @ R8=W2 + add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) + sub r3, r6, r11 @ R3=a0-W6*ROWr16[2] (a2) + mul r11, r8, r4 @ R11=W2*ROWr16[2] + sub r4, r6, r11 @ R4=a0-W2*ROWr16[2] (a3) + add r6, r6, r11 @ R6=a0+W2*ROWr16[2] (a0) - @@ at this point, R0=b0, R1=b1, R2=a1, R3=a2, R4=a3, - @@ R5=b2, R6=a0, R7=b3, R8=W2, R9=W4, R10=W6, R11 (free), - @@ R12=__const_ptr_, R14=&block[n] - @@ a0 += W4*row[4] - @@ a1 -= W4*row[4] - @@ a2 -= W4*row[4] - @@ a3 += W4*row[4] - ldrsh r11, [r14, #64] @ R11=ROWr16[4] - teq r11, #0 @ if null avoid muls - mulne r11, r9, r11 @ R11=W4*ROWr16[4] - @@ R9 is free now - addne r6, r6, r11 @ R6+=W4*ROWr16[4] (a0) - subne r2, r2, r11 @ R2-=W4*ROWr16[4] (a1) - subne r3, r3, r11 @ R3-=W4*ROWr16[4] (a2) - ldrsh r9, [r14, #96] @ R9=ROWr16[6] - addne r4, r4, r11 @ R4+=W4*ROWr16[4] (a3) - @@ W6 alone is no more useful, save W2*ROWr16[6] in it instead - teq r9, #0 @ if null avoid muls - mulne r11, r10, r9 @ R11=W6*ROWr16[6] - addne r6, r6, r11 @ R6+=W6*ROWr16[6] (a0) - mulne r10, r8, r9 @ R10=W2*ROWr16[6] - @@ a0 += W6*row[6]; - @@ a3 -= W6*row[6]; - @@ a1 -= W2*row[6]; - @@ a2 += W2*row[6]; - subne r4, r4, r11 @ R4-=W6*ROWr16[6] (a3) - subne r2, r2, r10 @ R2-=W2*ROWr16[6] (a1) - addne r3, r3, r10 @ R3+=W2*ROWr16[6] (a2) + @@ at this point, R0=b0, R1=b1, R2=a1, R3=a2, R4=a3, + @@ R5=b2, R6=a0, R7=b3, R8=W2, R9=W4, R10=W6, R11 (free), + @@ R12=__const_ptr_, R14=&block[n] + @@ a0 += W4*row[4] + @@ a1 -= W4*row[4] + @@ a2 -= W4*row[4] + @@ a3 += W4*row[4] + ldrsh r11, [r14, #64] @ R11=ROWr16[4] + teq r11, #0 @ if null avoid muls + mulne r11, r9, r11 @ R11=W4*ROWr16[4] + @@ R9 is free now + addne r6, r6, r11 @ R6+=W4*ROWr16[4] (a0) + subne r2, r2, r11 @ R2-=W4*ROWr16[4] (a1) + subne r3, r3, r11 @ R3-=W4*ROWr16[4] (a2) + ldrsh r9, [r14, #96] @ R9=ROWr16[6] + addne r4, r4, r11 @ R4+=W4*ROWr16[4] (a3) + @@ W6 alone is no more useful, save W2*ROWr16[6] in it instead + teq r9, #0 @ if null avoid muls + mulne r11, r10, r9 @ R11=W6*ROWr16[6] + addne r6, r6, r11 @ R6+=W6*ROWr16[6] (a0) + mulne r10, r8, r9 @ R10=W2*ROWr16[6] + @@ a0 += W6*row[6]; + @@ a3 -= W6*row[6]; + @@ a1 -= W2*row[6]; + @@ a2 += W2*row[6]; + subne r4, r4, r11 @ R4-=W6*ROWr16[6] (a3) + subne r2, r2, r10 @ R2-=W2*ROWr16[6] (a1) + addne r3, r3, r10 @ R3+=W2*ROWr16[6] (a2) __end_a_evaluation2: - @@ at this point, R0=b0, R1=b1, R2=a1, R3=a2, R4=a3, - @@ R5=b2, R6=a0, R7=b3, R8 (free), R9 (free), R10 (free), R11 (free), - @@ R12=__const_ptr_, R14=&block[n] - @@ col[0 ] = ((a0 + b0) >> COL_SHIFT); - @@ col[8 ] = ((a1 + b1) >> COL_SHIFT); - @@ col[16] = ((a2 + b2) >> COL_SHIFT); - @@ col[24] = ((a3 + b3) >> COL_SHIFT); - @@ col[32] = ((a3 - b3) >> COL_SHIFT); - @@ col[40] = ((a2 - b2) >> COL_SHIFT); - @@ col[48] = ((a1 - b1) >> COL_SHIFT); - @@ col[56] = ((a0 - b0) >> COL_SHIFT); - @@@@@ no optimisation here @@@@@ - add r8, r6, r0 @ R8=a0+b0 - add r9, r2, r1 @ R9=a1+b1 - mov r8, r8, asr #COL_SHIFT - mov r9, r9, asr #COL_SHIFT - strh r8, [r14, #0] - strh r9, [r14, #16] - add r8, r3, r5 @ R8=a2+b2 - add r9, r4, r7 @ R9=a3+b3 - mov r8, r8, asr #COL_SHIFT - mov r9, r9, asr #COL_SHIFT - strh r8, [r14, #32] - strh r9, [r14, #48] - sub r8, r4, r7 @ R8=a3-b3 - sub r9, r3, r5 @ R9=a2-b2 - mov r8, r8, asr #COL_SHIFT - mov r9, r9, asr #COL_SHIFT - strh r8, [r14, #64] - strh r9, [r14, #80] - sub r8, r2, r1 @ R8=a1-b1 - sub r9, r6, r0 @ R9=a0-b0 - mov r8, r8, asr #COL_SHIFT - mov r9, r9, asr #COL_SHIFT - strh r8, [r14, #96] - strh r9, [r14, #112] + @@ at this point, R0=b0, R1=b1, R2=a1, R3=a2, R4=a3, + @@ R5=b2, R6=a0, R7=b3, R8 (free), R9 (free), R10 (free), R11 (free), + @@ R12=__const_ptr_, R14=&block[n] + @@ col[0 ] = ((a0 + b0) >> COL_SHIFT); + @@ col[8 ] = ((a1 + b1) >> COL_SHIFT); + @@ col[16] = ((a2 + b2) >> COL_SHIFT); + @@ col[24] = ((a3 + b3) >> COL_SHIFT); + @@ col[32] = ((a3 - b3) >> COL_SHIFT); + @@ col[40] = ((a2 - b2) >> COL_SHIFT); + @@ col[48] = ((a1 - b1) >> COL_SHIFT); + @@ col[56] = ((a0 - b0) >> COL_SHIFT); + @@@@@ no optimisation here @@@@@ + add r8, r6, r0 @ R8=a0+b0 + add r9, r2, r1 @ R9=a1+b1 + mov r8, r8, asr #COL_SHIFT + mov r9, r9, asr #COL_SHIFT + strh r8, [r14, #0] + strh r9, [r14, #16] + add r8, r3, r5 @ R8=a2+b2 + add r9, r4, r7 @ R9=a3+b3 + mov r8, r8, asr #COL_SHIFT + mov r9, r9, asr #COL_SHIFT + strh r8, [r14, #32] + strh r9, [r14, #48] + sub r8, r4, r7 @ R8=a3-b3 + sub r9, r3, r5 @ R9=a2-b2 + mov r8, r8, asr #COL_SHIFT + mov r9, r9, asr #COL_SHIFT + strh r8, [r14, #64] + strh r9, [r14, #80] + sub r8, r2, r1 @ R8=a1-b1 + sub r9, r6, r0 @ R9=a0-b0 + mov r8, r8, asr #COL_SHIFT + mov r9, r9, asr #COL_SHIFT + strh r8, [r14, #96] + strh r9, [r14, #112] __end_col_loop: - @@ at this point, R0-R11 (free) - @@ R12=__const_ptr_, R14=&block[n] - ldr r0, [sp, #0] @ R0=block - teq r0, r14 @ compare current &block[n] to block, when block is reached, the loop is finished. - sub r14, r14, #2 - bne __col_loop + @@ at this point, R0-R11 (free) + @@ R12=__const_ptr_, R14=&block[n] + ldr r0, [sp, #0] @ R0=block + teq r0, r14 @ compare current &block[n] to block, when block is reached, the loop is finished. + sub r14, r14, #2 + bne __col_loop @@ -466,15 +466,15 @@ @@ kind of sub-function, here not to overload the common case. __end_bef_a_evaluation: - add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) + add r2, r6, r11 @ R2=a0+W6*ROWr16[2] (a1) mul r11, r8, r4 @ R11=W2*ROWr16[2] sub r4, r6, r11 @ R4=a0-W2*ROWr16[2] (a3) add r6, r6, r11 @ R6=a0+W2*ROWr16[2] (a0) - bal __end_a_evaluation + bal __end_a_evaluation __constant_ptr__: @@ see #defines at the beginning of the source code for values. - .align + .align .word W1 .word W2 .word W3
--- a/avcodec.h Wed Dec 21 17:50:40 2005 +0000 +++ b/avcodec.h Thu Dec 22 01:10:11 2005 +0000 @@ -15,21 +15,21 @@ #include <sys/types.h> /* size_t */ //FIXME the following 2 really dont belong in here -#define FFMPEG_VERSION_INT 0x000409 -#define FFMPEG_VERSION "CVS" +#define FFMPEG_VERSION_INT 0x000409 +#define FFMPEG_VERSION "CVS" -#define AV_STRINGIFY(s) AV_TOSTRING(s) +#define AV_STRINGIFY(s) AV_TOSTRING(s) #define AV_TOSTRING(s) #s -#define LIBAVCODEC_VERSION_INT ((51<<16)+(0<<8)+0) -#define LIBAVCODEC_VERSION 51.0.0 -#define LIBAVCODEC_BUILD LIBAVCODEC_VERSION_INT +#define LIBAVCODEC_VERSION_INT ((51<<16)+(0<<8)+0) +#define LIBAVCODEC_VERSION 51.0.0 +#define LIBAVCODEC_BUILD LIBAVCODEC_VERSION_INT -#define LIBAVCODEC_IDENT "Lavc" AV_STRINGIFY(LIBAVCODEC_VERSION) +#define LIBAVCODEC_IDENT "Lavc" AV_STRINGIFY(LIBAVCODEC_VERSION) -#define AV_NOPTS_VALUE int64_t_C(0x8000000000000000) -#define AV_TIME_BASE 1000000 -#define AV_TIME_BASE_Q (AVRational){1, AV_TIME_BASE} +#define AV_NOPTS_VALUE int64_t_C(0x8000000000000000) +#define AV_TIME_BASE 1000000 +#define AV_TIME_BASE_Q (AVRational){1, AV_TIME_BASE} enum CodecID { CODEC_ID_NONE, @@ -362,9 +362,9 @@ #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< place global headers at every keyframe instead of in extradata /* Unsupported options : - * Syntax Arithmetic coding (SAC) - * Reference Picture Selection - * Independant Segment Decoding */ + * Syntax Arithmetic coding (SAC) + * Reference Picture Selection + * Independant Segment Decoding */ /* /Fx */ /* codec capabilities */ @@ -646,9 +646,9 @@ */\ int8_t *ref_index[2]; -#define FF_QSCALE_TYPE_MPEG1 0 -#define FF_QSCALE_TYPE_MPEG2 1 -#define FF_QSCALE_TYPE_H264 2 +#define FF_QSCALE_TYPE_MPEG1 0 +#define FF_QSCALE_TYPE_MPEG2 1 +#define FF_QSCALE_TYPE_H264 2 #define FF_BUFFER_TYPE_INTERNAL 1 #define FF_BUFFER_TYPE_USER 2 ///< Direct rendering buffers (image is (de)allocated by user) @@ -684,9 +684,9 @@ struct AVCLASS { const char* class_name; const char* (*item_name)(void*); /* actually passing a pointer to an AVCodecContext - or AVFormatContext, which begin with an AVClass. - Needed because av_log is in libavcodec and has no visibility - of AVIn/OutputFormat */ + or AVFormatContext, which begin with an AVClass. + Needed because av_log is in libavcodec and has no visibility + of AVIn/OutputFormat */ struct AVOption *option; }; @@ -1252,18 +1252,18 @@ * result into program crash) */ unsigned dsp_mask; -#define FF_MM_FORCE 0x80000000 /* force usage of selected flags (OR) */ +#define FF_MM_FORCE 0x80000000 /* force usage of selected flags (OR) */ /* lower 16 bits - CPU features */ #ifdef HAVE_MMX -#define FF_MM_MMX 0x0001 /* standard MMX */ -#define FF_MM_3DNOW 0x0004 /* AMD 3DNOW */ -#define FF_MM_MMXEXT 0x0002 /* SSE integer functions or AMD MMX ext */ -#define FF_MM_SSE 0x0008 /* SSE functions */ -#define FF_MM_SSE2 0x0010 /* PIV SSE2 functions */ -#define FF_MM_3DNOWEXT 0x0020 /* AMD 3DNowExt */ +#define FF_MM_MMX 0x0001 /* standard MMX */ +#define FF_MM_3DNOW 0x0004 /* AMD 3DNOW */ +#define FF_MM_MMXEXT 0x0002 /* SSE integer functions or AMD MMX ext */ +#define FF_MM_SSE 0x0008 /* SSE functions */ +#define FF_MM_SSE2 0x0010 /* PIV SSE2 functions */ +#define FF_MM_3DNOWEXT 0x0020 /* AMD 3DNowExt */ #endif /* HAVE_MMX */ #ifdef HAVE_IWMMXT -#define FF_MM_IWMMXT 0x0100 /* XScale IWMMXT */ +#define FF_MM_IWMMXT 0x0100 /* XScale IWMMXT */ #endif /* HAVE_IWMMXT */ /** @@ -2223,7 +2223,7 @@ #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */ #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */ int img_get_alpha_info(const AVPicture *src, - int pix_fmt, int width, int height); + int pix_fmt, int width, int height); /* convert among pixel formats */ int img_convert(AVPicture *dst, int dst_pix_fmt,
--- a/beosthread.c Wed Dec 21 17:50:40 2005 +0000 +++ b/beosthread.c Thu Dec 22 01:10:11 2005 +0000 @@ -35,20 +35,20 @@ // it's odd Be never patented that :D struct benaphore { - vint32 atom; - sem_id sem; + vint32 atom; + sem_id sem; }; static inline int lock_ben(struct benaphore *ben) { - if (atomic_add(&ben->atom, 1) > 0) - return acquire_sem(ben->sem); - return B_OK; + if (atomic_add(&ben->atom, 1) > 0) + return acquire_sem(ben->sem); + return B_OK; } static inline int unlock_ben(struct benaphore *ben) { - if (atomic_add(&ben->atom, -1) > 1) - return release_sem(ben->sem); - return B_OK; + if (atomic_add(&ben->atom, -1) > 1) + return release_sem(ben->sem); + return B_OK; } static struct benaphore av_thread_lib_ben; @@ -155,25 +155,25 @@ int avcodec_thread_lock_lib(void) { - return lock_ben(&av_thread_lib_ben); + return lock_ben(&av_thread_lib_ben); } int avcodec_thread_unlock_lib(void) { - return unlock_ben(&av_thread_lib_ben); + return unlock_ben(&av_thread_lib_ben); } /* our versions of _init and _fini (which are called by those actually from crt.o) */ void initialize_after(void) { - av_thread_lib_ben.atom = 0; - av_thread_lib_ben.sem = create_sem(0, "libavcodec benaphore"); + av_thread_lib_ben.atom = 0; + av_thread_lib_ben.sem = create_sem(0, "libavcodec benaphore"); } void uninitialize_before(void) { - delete_sem(av_thread_lib_ben.sem); + delete_sem(av_thread_lib_ben.sem); }
--- a/bitstream.c Wed Dec 21 17:50:40 2005 +0000 +++ b/bitstream.c Thu Dec 22 01:10:11 2005 +0000 @@ -83,7 +83,7 @@ { int bit= get_bits1(s); if(!bit) - av_log(NULL, AV_LOG_INFO, "Marker bit missing %s\n", msg); + av_log(NULL, AV_LOG_INFO, "Marker bit missing %s\n", msg); return bit; }
--- a/bitstream.h Wed Dec 21 17:50:40 2005 +0000 +++ b/bitstream.h Thu Dec 22 01:10:11 2005 +0000 @@ -146,7 +146,7 @@ # ifdef __GNUC__ static inline uint32_t unaligned32(const void *v) { struct Unaligned { - uint32_t i; + uint32_t i; } __attribute__((packed)); return ((const struct Unaligned *) v)->i; @@ -183,7 +183,7 @@ bit_buf = (bit_buf<<n) | value; bit_left-=n; } else { - bit_buf<<=bit_left; + bit_buf<<=bit_left; bit_buf |= value >> (n - bit_left); #ifdef UNALIGNED_STORES_ARE_BAD if (3 & (intptr_t) s->buf_ptr) { @@ -196,7 +196,7 @@ *(uint32_t *)s->buf_ptr = be2me_32(bit_buf); //printf("bitbuf = %08x\n", bit_buf); s->buf_ptr+=4; - bit_left+=32 - n; + bit_left+=32 - n; bit_buf = value; } @@ -212,21 +212,21 @@ # ifdef ALIGNED_BITSTREAM_WRITER # if defined(ARCH_X86) || defined(ARCH_X86_64) asm volatile( - "movl %0, %%ecx \n\t" - "xorl %%eax, %%eax \n\t" - "shrdl %%cl, %1, %%eax \n\t" - "shrl %%cl, %1 \n\t" - "movl %0, %%ecx \n\t" - "shrl $3, %%ecx \n\t" - "andl $0xFFFFFFFC, %%ecx \n\t" - "bswapl %1 \n\t" - "orl %1, (%2, %%ecx) \n\t" - "bswapl %%eax \n\t" - "addl %3, %0 \n\t" - "movl %%eax, 4(%2, %%ecx) \n\t" - : "=&r" (s->index), "=&r" (value) - : "r" (s->buf), "r" (n), "0" (s->index), "1" (value<<(-n)) - : "%eax", "%ecx" + "movl %0, %%ecx \n\t" + "xorl %%eax, %%eax \n\t" + "shrdl %%cl, %1, %%eax \n\t" + "shrl %%cl, %1 \n\t" + "movl %0, %%ecx \n\t" + "shrl $3, %%ecx \n\t" + "andl $0xFFFFFFFC, %%ecx \n\t" + "bswapl %1 \n\t" + "orl %1, (%2, %%ecx) \n\t" + "bswapl %%eax \n\t" + "addl %3, %0 \n\t" + "movl %%eax, 4(%2, %%ecx) \n\t" + : "=&r" (s->index), "=&r" (value) + : "r" (s->buf), "r" (n), "0" (s->index), "1" (value<<(-n)) + : "%eax", "%ecx" ); # else int index= s->index; @@ -243,20 +243,20 @@ # else //ALIGNED_BITSTREAM_WRITER # if defined(ARCH_X86) || defined(ARCH_X86_64) asm volatile( - "movl $7, %%ecx \n\t" - "andl %0, %%ecx \n\t" - "addl %3, %%ecx \n\t" - "negl %%ecx \n\t" - "shll %%cl, %1 \n\t" - "bswapl %1 \n\t" - "movl %0, %%ecx \n\t" - "shrl $3, %%ecx \n\t" - "orl %1, (%%ecx, %2) \n\t" - "addl %3, %0 \n\t" - "movl $0, 4(%%ecx, %2) \n\t" - : "=&r" (s->index), "=&r" (value) - : "r" (s->buf), "r" (n), "0" (s->index), "1" (value) - : "%ecx" + "movl $7, %%ecx \n\t" + "andl %0, %%ecx \n\t" + "addl %3, %%ecx \n\t" + "negl %%ecx \n\t" + "shll %%cl, %1 \n\t" + "bswapl %1 \n\t" + "movl %0, %%ecx \n\t" + "shrl $3, %%ecx \n\t" + "orl %1, (%%ecx, %2) \n\t" + "addl %3, %0 \n\t" + "movl $0, 4(%%ecx, %2) \n\t" + : "=&r" (s->index), "=&r" (value) + : "r" (s->buf), "r" (n), "0" (s->index), "1" (value) + : "%ecx" ); # else int index= s->index; @@ -276,9 +276,9 @@ static inline uint8_t* pbBufPtr(PutBitContext *s) { #ifdef ALT_BITSTREAM_WRITER - return s->buf + (s->index>>3); + return s->buf + (s->index>>3); #else - return s->buf_ptr; + return s->buf_ptr; #endif } @@ -290,10 +290,10 @@ assert((put_bits_count(s)&7)==0); #ifdef ALT_BITSTREAM_WRITER FIXME may need some cleaning of the buffer - s->index += n<<3; + s->index += n<<3; #else assert(s->bit_left==32); - s->buf_ptr += n; + s->buf_ptr += n; #endif } @@ -366,10 +366,10 @@ static inline int unaligned32_be(const void *v) { #ifdef CONFIG_ALIGN - const uint8_t *p=v; - return (((p[0]<<8) | p[1])<<16) | (p[2]<<8) | (p[3]); + const uint8_t *p=v; + return (((p[0]<<8) | p[1])<<16) | (p[2]<<8) | (p[3]); #else - return be2me_32( unaligned32(v)); //original + return be2me_32( unaligned32(v)); //original #endif } @@ -528,8 +528,8 @@ #if defined(ARCH_X86) || defined(ARCH_X86_64) # define SKIP_CACHE(name, gb, num)\ asm(\ - "shldl %2, %1, %0 \n\t"\ - "shll %2, %1 \n\t"\ + "shldl %2, %1, %0 \n\t"\ + "shll %2, %1 \n\t"\ : "+r" (name##_cache0), "+r" (name##_cache1)\ : "Ic" ((uint8_t)num)\ );
--- a/cljr.c Wed Dec 21 17:50:40 2005 +0000 +++ b/cljr.c Thu Dec 22 01:10:11 2005 +0000 @@ -61,13 +61,13 @@ uint8_t *cb= &a->picture.data[1][ y*a->picture.linesize[1] ]; uint8_t *cr= &a->picture.data[2][ y*a->picture.linesize[2] ]; for(x=0; x<avctx->width; x+=4){ - luma[3] = get_bits(&a->gb, 5) << 3; - luma[2] = get_bits(&a->gb, 5) << 3; - luma[1] = get_bits(&a->gb, 5) << 3; - luma[0] = get_bits(&a->gb, 5) << 3; - luma+= 4; - *(cb++) = get_bits(&a->gb, 6) << 2; - *(cr++) = get_bits(&a->gb, 6) << 2; + luma[3] = get_bits(&a->gb, 5) << 3; + luma[2] = get_bits(&a->gb, 5) << 3; + luma[1] = get_bits(&a->gb, 5) << 3; + luma[0] = get_bits(&a->gb, 5) << 3; + luma+= 4; + *(cb++) = get_bits(&a->gb, 6) << 2; + *(cr++) = get_bits(&a->gb, 6) << 2; } }
--- a/dct-test.c Wed Dec 21 17:50:40 2005 +0000 +++ b/dct-test.c Thu Dec 22 01:10:11 2005 +0000 @@ -65,14 +65,14 @@ static short idct_mmx_perm[64]; static short idct_simple_mmx_perm[64]={ - 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, - 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, - 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, - 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, - 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, - 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, - 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, - 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, + 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, + 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, + 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, + 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, + 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, + 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, + 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, + 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, }; void idct_mmx_init(void) @@ -81,8 +81,8 @@ /* the mmx/mmxext idct uses a reordered input, so we patch scan tables */ for (i = 0; i < 64; i++) { - idct_mmx_perm[i] = (i & 0x38) | ((i & 6) >> 1) | ((i & 1) << 2); -// idct_simple_mmx_perm[i] = simple_block_permute_op(i); + idct_mmx_perm[i] = (i & 0x38) | ((i & 6) >> 1) | ((i & 1) << 2); +// idct_simple_mmx_perm[i] = simple_block_permute_op(i); } } @@ -151,7 +151,7 @@ for(i=0;i<64;i++) block[idct_simple_mmx_perm[i]] = block1[i]; - } else { + } else { for(i=0; i<64; i++) block[i]= block1[i]; } @@ -186,9 +186,9 @@ if (v > err_inf) err_inf = v; err2 += v * v; - sysErr[i] += block[i] - block1[i]; - blockSumErr += v; - if( abs(block[i])>maxout) maxout=abs(block[i]); + sysErr[i] += block[i] - block1[i]; + blockSumErr += v; + if( abs(block[i])>maxout) maxout=abs(block[i]); } if(blockSumErrMax < blockSumErr) blockSumErrMax= blockSumErr; #if 0 // print different matrix pairs @@ -209,7 +209,7 @@ #if 1 // dump systematic errors for(i=0; i<64; i++){ - if(i%8==0) printf("\n"); + if(i%8==0) printf("\n"); printf("%5d ", (int)sysErr[i]); } printf("\n"); @@ -503,7 +503,7 @@ dct_error("XVID-MMX2", 1, ff_idct_xvid_mmx2, idct, test); // dct_error("ODIVX-C", 1, odivx_idct_c, idct); //printf(" test against odivx idct\n"); - // dct_error("REF", 1, idct, odivx_idct_c); + // dct_error("REF", 1, idct, odivx_idct_c); // dct_error("INT", 1, j_rev_dct, odivx_idct_c); // dct_error("MMX", 1, ff_mmx_idct, odivx_idct_c); // dct_error("MMXEXT", 1, ff_mmxext_idct, odivx_idct_c);
--- a/dsputil.c Wed Dec 21 17:50:40 2005 +0000 +++ b/dsputil.c Thu Dec 22 01:10:11 2005 +0000 @@ -124,14 +124,14 @@ /* Input permutation for the simple_idct_mmx */ static const uint8_t simple_mmx_permutation[64]={ - 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, - 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, - 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, - 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, - 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, - 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, - 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, - 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, + 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, + 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, + 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, + 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, + 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, + 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, + 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, + 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, }; static int pix_sum_c(uint8_t * pix, int line_size) @@ -140,18 +140,18 @@ s = 0; for (i = 0; i < 16; i++) { - for (j = 0; j < 16; j += 8) { - s += pix[0]; - s += pix[1]; - s += pix[2]; - s += pix[3]; - s += pix[4]; - s += pix[5]; - s += pix[6]; - s += pix[7]; - pix += 8; - } - pix += line_size - 16; + for (j = 0; j < 16; j += 8) { + s += pix[0]; + s += pix[1]; + s += pix[2]; + s += pix[3]; + s += pix[4]; + s += pix[5]; + s += pix[6]; + s += pix[7]; + pix += 8; + } + pix += line_size - 16; } return s; } @@ -163,33 +163,33 @@ s = 0; for (i = 0; i < 16; i++) { - for (j = 0; j < 16; j += 8) { + for (j = 0; j < 16; j += 8) { #if 0 - s += sq[pix[0]]; - s += sq[pix[1]]; - s += sq[pix[2]]; - s += sq[pix[3]]; - s += sq[pix[4]]; - s += sq[pix[5]]; - s += sq[pix[6]]; - s += sq[pix[7]]; + s += sq[pix[0]]; + s += sq[pix[1]]; + s += sq[pix[2]]; + s += sq[pix[3]]; + s += sq[pix[4]]; + s += sq[pix[5]]; + s += sq[pix[6]]; + s += sq[pix[7]]; #else #if LONG_MAX > 2147483647 - register uint64_t x=*(uint64_t*)pix; - s += sq[x&0xff]; - s += sq[(x>>8)&0xff]; - s += sq[(x>>16)&0xff]; - s += sq[(x>>24)&0xff]; + register uint64_t x=*(uint64_t*)pix; + s += sq[x&0xff]; + s += sq[(x>>8)&0xff]; + s += sq[(x>>16)&0xff]; + s += sq[(x>>24)&0xff]; s += sq[(x>>32)&0xff]; s += sq[(x>>40)&0xff]; s += sq[(x>>48)&0xff]; s += sq[(x>>56)&0xff]; #else - register uint32_t x=*(uint32_t*)pix; - s += sq[x&0xff]; - s += sq[(x>>8)&0xff]; - s += sq[(x>>16)&0xff]; - s += sq[(x>>24)&0xff]; + register uint32_t x=*(uint32_t*)pix; + s += sq[x&0xff]; + s += sq[(x>>8)&0xff]; + s += sq[(x>>16)&0xff]; + s += sq[(x>>24)&0xff]; x=*(uint32_t*)(pix+4); s += sq[x&0xff]; s += sq[(x>>8)&0xff]; @@ -197,9 +197,9 @@ s += sq[(x>>24)&0xff]; #endif #endif - pix += 8; - } - pix += line_size - 16; + pix += 8; + } + pix += line_size - 16; } return s; } @@ -410,7 +410,7 @@ } static void diff_pixels_c(DCTELEM *restrict block, const uint8_t *s1, - const uint8_t *s2, int stride){ + const uint8_t *s2, int stride){ int i; /* read the pixels */ @@ -431,7 +431,7 @@ static void put_pixels_clamped_c(const DCTELEM *block, uint8_t *restrict pixels, - int line_size) + int line_size) { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; @@ -453,7 +453,7 @@ } static void put_pixels_clamped4_c(const DCTELEM *block, uint8_t *restrict pixels, - int line_size) + int line_size) { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; @@ -471,7 +471,7 @@ } static void put_pixels_clamped2_c(const DCTELEM *block, uint8_t *restrict pixels, - int line_size) + int line_size) { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; @@ -1214,7 +1214,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (683*(2*src[j] + src[j+1] + 1)) >> 11; + dst[j] = (683*(2*src[j] + src[j+1] + 1)) >> 11; } src += stride; dst += stride; @@ -1225,7 +1225,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (683*(src[j] + 2*src[j+1] + 1)) >> 11; + dst[j] = (683*(src[j] + 2*src[j+1] + 1)) >> 11; } src += stride; dst += stride; @@ -1236,7 +1236,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (683*(2*src[j] + src[j+stride] + 1)) >> 11; + dst[j] = (683*(2*src[j] + src[j+stride] + 1)) >> 11; } src += stride; dst += stride; @@ -1247,7 +1247,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15; + dst[j] = (2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15; } src += stride; dst += stride; @@ -1258,7 +1258,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15; + dst[j] = (2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15; } src += stride; dst += stride; @@ -1269,7 +1269,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (683*(src[j] + 2*src[j+stride] + 1)) >> 11; + dst[j] = (683*(src[j] + 2*src[j+stride] + 1)) >> 11; } src += stride; dst += stride; @@ -1280,7 +1280,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15; + dst[j] = (2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15; } src += stride; dst += stride; @@ -1291,7 +1291,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15; + dst[j] = (2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15; } src += stride; dst += stride; @@ -1311,7 +1311,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((683*(2*src[j] + src[j+1] + 1)) >> 11) + 1) >> 1; + dst[j] = (dst[j] + ((683*(2*src[j] + src[j+1] + 1)) >> 11) + 1) >> 1; } src += stride; dst += stride; @@ -1322,7 +1322,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+1] + 1)) >> 11) + 1) >> 1; + dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+1] + 1)) >> 11) + 1) >> 1; } src += stride; dst += stride; @@ -1333,7 +1333,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((683*(2*src[j] + src[j+stride] + 1)) >> 11) + 1) >> 1; + dst[j] = (dst[j] + ((683*(2*src[j] + src[j+stride] + 1)) >> 11) + 1) >> 1; } src += stride; dst += stride; @@ -1344,7 +1344,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15) + 1) >> 1; + dst[j] = (dst[j] + ((2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15) + 1) >> 1; } src += stride; dst += stride; @@ -1355,7 +1355,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1; + dst[j] = (dst[j] + ((2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1; } src += stride; dst += stride; @@ -1366,7 +1366,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+stride] + 1)) >> 11) + 1) >> 1; + dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+stride] + 1)) >> 11) + 1) >> 1; } src += stride; dst += stride; @@ -1377,7 +1377,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1; + dst[j] = (dst[j] + ((2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1; } src += stride; dst += stride; @@ -1388,7 +1388,7 @@ int i,j; for (i=0; i < height; i++) { for (j=0; j < width; j++) { - dst[j] = (dst[j] + ((2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15) + 1) >> 1; + dst[j] = (dst[j] + ((2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15) + 1) >> 1; } src += stride; dst += stride; @@ -3666,15 +3666,15 @@ #ifdef CONFIG_ENCODERS if(avctx->dct_algo==FF_DCT_FASTINT) { c->fdct = fdct_ifast; - c->fdct248 = fdct_ifast248; + c->fdct248 = fdct_ifast248; } else if(avctx->dct_algo==FF_DCT_FAAN) { c->fdct = ff_faandct; - c->fdct248 = ff_faandct248; + c->fdct248 = ff_faandct248; } else { c->fdct = ff_jpeg_fdct_islow; //slow/accurate/default - c->fdct248 = ff_fdct248_islow; + c->fdct248 = ff_fdct248_islow; } #endif //CONFIG_ENCODERS
--- a/dsputil.h Wed Dec 21 17:50:40 2005 +0000 +++ b/dsputil.h Thu Dec 22 01:10:11 2005 +0000 @@ -151,7 +151,7 @@ * global motion compensation. */ void (*gmc )(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int ox, int oy, - int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height); + int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height); void (*clear_blocks)(DCTELEM *blocks/*align 16*/); int (*pix_sum)(uint8_t * pix, int line_size); int (*pix_norm1)(uint8_t * pix, int line_size); @@ -342,7 +342,7 @@ void ff_set_cmp(DSPContext* c, me_cmp_func *cmp, int type); -#define BYTE_VEC32(c) ((c)*0x01010101UL) +#define BYTE_VEC32(c) ((c)*0x01010101UL) static inline uint32_t rnd_avg32(uint32_t a, uint32_t b) {
--- a/dtsdec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/dtsdec.c Thu Dec 22 01:10:11 2005 +0000 @@ -194,7 +194,7 @@ { if (flags & DTS_LFE) return 6; - else if (flags & 1) /* center channel */ + else if (flags & 1) /* center channel */ return 5; else if ((flags & DTS_CHANNEL_MASK) == DTS_2F2R) return 4;
--- a/dv.c Wed Dec 21 17:50:40 2005 +0000 +++ b/dv.c Thu Dec 22 01:10:11 2005 +0000 @@ -84,7 +84,7 @@ j = perm[i]; s->dv_idct_shift[0][0][q][j] = dv_quant_shifts[q][dv_88_areas[i]] + 1; - s->dv_idct_shift[1][0][q][j] = s->dv_idct_shift[0][0][q][j] + 1; + s->dv_idct_shift[1][0][q][j] = s->dv_idct_shift[0][0][q][j] + 1; } /* 248DCT */ @@ -92,7 +92,7 @@ /* 248 table */ s->dv_idct_shift[0][1][q][i] = dv_quant_shifts[q][dv_248_areas[i]] + 1; - s->dv_idct_shift[1][1][q][i] = s->dv_idct_shift[0][1][q][i] + 1; + s->dv_idct_shift[1][1][q][i] = s->dv_idct_shift[0][1][q][i] + 1; } } } @@ -114,35 +114,35 @@ done = 1; dv_vlc_map = av_mallocz_static(DV_VLC_MAP_LEV_SIZE*DV_VLC_MAP_RUN_SIZE*sizeof(struct dv_vlc_pair)); - if (!dv_vlc_map) - return -ENOMEM; + if (!dv_vlc_map) + return -ENOMEM; - /* dv_anchor lets each thread know its Id */ - dv_anchor = av_malloc(12*27*sizeof(void*)); - if (!dv_anchor) { - return -ENOMEM; - } - for (i=0; i<12*27; i++) - dv_anchor[i] = (void*)(size_t)i; + /* dv_anchor lets each thread know its Id */ + dv_anchor = av_malloc(12*27*sizeof(void*)); + if (!dv_anchor) { + return -ENOMEM; + } + for (i=0; i<12*27; i++) + dv_anchor[i] = (void*)(size_t)i; - /* it's faster to include sign bit in a generic VLC parsing scheme */ - for (i=0, j=0; i<NB_DV_VLC; i++, j++) { - new_dv_vlc_bits[j] = dv_vlc_bits[i]; - new_dv_vlc_len[j] = dv_vlc_len[i]; - new_dv_vlc_run[j] = dv_vlc_run[i]; - new_dv_vlc_level[j] = dv_vlc_level[i]; + /* it's faster to include sign bit in a generic VLC parsing scheme */ + for (i=0, j=0; i<NB_DV_VLC; i++, j++) { + new_dv_vlc_bits[j] = dv_vlc_bits[i]; + new_dv_vlc_len[j] = dv_vlc_len[i]; + new_dv_vlc_run[j] = dv_vlc_run[i]; + new_dv_vlc_level[j] = dv_vlc_level[i]; - if (dv_vlc_level[i]) { - new_dv_vlc_bits[j] <<= 1; - new_dv_vlc_len[j]++; + if (dv_vlc_level[i]) { + new_dv_vlc_bits[j] <<= 1; + new_dv_vlc_len[j]++; - j++; - new_dv_vlc_bits[j] = (dv_vlc_bits[i] << 1) | 1; - new_dv_vlc_len[j] = dv_vlc_len[i] + 1; - new_dv_vlc_run[j] = dv_vlc_run[i]; - new_dv_vlc_level[j] = -dv_vlc_level[i]; - } - } + j++; + new_dv_vlc_bits[j] = (dv_vlc_bits[i] << 1) | 1; + new_dv_vlc_len[j] = dv_vlc_len[i] + 1; + new_dv_vlc_run[j] = dv_vlc_run[i]; + new_dv_vlc_level[j] = -dv_vlc_level[i]; + } + } /* NOTE: as a trick, we use the fact the no codes are unused to accelerate the parsing of partial codes */ @@ -150,10 +150,10 @@ new_dv_vlc_len, 1, 1, new_dv_vlc_bits, 2, 2, 0); dv_rl_vlc = av_malloc(dv_vlc.table_size * sizeof(RL_VLC_ELEM)); - if (!dv_rl_vlc) { - av_free(dv_anchor); - return -ENOMEM; - } + if (!dv_rl_vlc) { + av_free(dv_anchor); + return -ENOMEM; + } for(i = 0; i < dv_vlc.table_size; i++){ int code= dv_vlc.table[i][0]; int len = dv_vlc.table[i][1]; @@ -170,49 +170,49 @@ dv_rl_vlc[i].level = level; dv_rl_vlc[i].run = run; } - free_vlc(&dv_vlc); + free_vlc(&dv_vlc); - for (i = 0; i < NB_DV_VLC - 1; i++) { + for (i = 0; i < NB_DV_VLC - 1; i++) { if (dv_vlc_run[i] >= DV_VLC_MAP_RUN_SIZE) - continue; + continue; #ifdef DV_CODEC_TINY_TARGET if (dv_vlc_level[i] >= DV_VLC_MAP_LEV_SIZE) - continue; + continue; #endif - if (dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].size != 0) - continue; + if (dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].size != 0) + continue; - dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].vlc = dv_vlc_bits[i] << - (!!dv_vlc_level[i]); - dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].size = dv_vlc_len[i] + - (!!dv_vlc_level[i]); - } - for (i = 0; i < DV_VLC_MAP_RUN_SIZE; i++) { + dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].vlc = dv_vlc_bits[i] << + (!!dv_vlc_level[i]); + dv_vlc_map[dv_vlc_run[i]][dv_vlc_level[i]].size = dv_vlc_len[i] + + (!!dv_vlc_level[i]); + } + for (i = 0; i < DV_VLC_MAP_RUN_SIZE; i++) { #ifdef DV_CODEC_TINY_TARGET - for (j = 1; j < DV_VLC_MAP_LEV_SIZE; j++) { - if (dv_vlc_map[i][j].size == 0) { - dv_vlc_map[i][j].vlc = dv_vlc_map[0][j].vlc | - (dv_vlc_map[i-1][0].vlc << (dv_vlc_map[0][j].size)); - dv_vlc_map[i][j].size = dv_vlc_map[i-1][0].size + - dv_vlc_map[0][j].size; - } - } + for (j = 1; j < DV_VLC_MAP_LEV_SIZE; j++) { + if (dv_vlc_map[i][j].size == 0) { + dv_vlc_map[i][j].vlc = dv_vlc_map[0][j].vlc | + (dv_vlc_map[i-1][0].vlc << (dv_vlc_map[0][j].size)); + dv_vlc_map[i][j].size = dv_vlc_map[i-1][0].size + + dv_vlc_map[0][j].size; + } + } #else - for (j = 1; j < DV_VLC_MAP_LEV_SIZE/2; j++) { - if (dv_vlc_map[i][j].size == 0) { - dv_vlc_map[i][j].vlc = dv_vlc_map[0][j].vlc | - (dv_vlc_map[i-1][0].vlc << (dv_vlc_map[0][j].size)); - dv_vlc_map[i][j].size = dv_vlc_map[i-1][0].size + - dv_vlc_map[0][j].size; - } - dv_vlc_map[i][((uint16_t)(-j))&0x1ff].vlc = - dv_vlc_map[i][j].vlc | 1; - dv_vlc_map[i][((uint16_t)(-j))&0x1ff].size = - dv_vlc_map[i][j].size; - } + for (j = 1; j < DV_VLC_MAP_LEV_SIZE/2; j++) { + if (dv_vlc_map[i][j].size == 0) { + dv_vlc_map[i][j].vlc = dv_vlc_map[0][j].vlc | + (dv_vlc_map[i-1][0].vlc << (dv_vlc_map[0][j].size)); + dv_vlc_map[i][j].size = dv_vlc_map[i-1][0].size + + dv_vlc_map[0][j].size; + } + dv_vlc_map[i][((uint16_t)(-j))&0x1ff].vlc = + dv_vlc_map[i][j].vlc | 1; + dv_vlc_map[i][((uint16_t)(-j))&0x1ff].size = + dv_vlc_map[i][j].size; + } #endif - } + } } /* Generic DSP setup */ @@ -241,7 +241,7 @@ /* FIXME: I really don't think this should be here */ if (dv_codec_profile(avctx)) - avctx->pix_fmt = dv_codec_profile(avctx)->pix_fmt; + avctx->pix_fmt = dv_codec_profile(avctx)->pix_fmt; avctx->coded_frame = &s->picture; s->avctx= avctx; @@ -306,9 +306,9 @@ /* if we must parse a partial vlc, we do it here */ if (partial_bit_count > 0) { re_cache = ((unsigned)re_cache >> partial_bit_count) | - (mb->partial_bit_buffer << (sizeof(re_cache)*8 - partial_bit_count)); - re_index -= partial_bit_count; - mb->partial_bit_count = 0; + (mb->partial_bit_buffer << (sizeof(re_cache)*8 - partial_bit_count)); + re_index -= partial_bit_count; + mb->partial_bit_count = 0; } /* get the AC coefficients until last_index is reached */ @@ -318,30 +318,30 @@ #endif /* our own optimized GET_RL_VLC */ index = NEG_USR32(re_cache, TEX_VLC_BITS); - vlc_len = dv_rl_vlc[index].len; + vlc_len = dv_rl_vlc[index].len; if (vlc_len < 0) { index = NEG_USR32((unsigned)re_cache << TEX_VLC_BITS, -vlc_len) + dv_rl_vlc[index].level; vlc_len = TEX_VLC_BITS - vlc_len; } level = dv_rl_vlc[index].level; - run = dv_rl_vlc[index].run; + run = dv_rl_vlc[index].run; - /* gotta check if we're still within gb boundaries */ - if (re_index + vlc_len > last_index) { - /* should be < 16 bits otherwise a codeword could have been parsed */ - mb->partial_bit_count = last_index - re_index; - mb->partial_bit_buffer = NEG_USR32(re_cache, mb->partial_bit_count); - re_index = last_index; - break; - } - re_index += vlc_len; + /* gotta check if we're still within gb boundaries */ + if (re_index + vlc_len > last_index) { + /* should be < 16 bits otherwise a codeword could have been parsed */ + mb->partial_bit_count = last_index - re_index; + mb->partial_bit_buffer = NEG_USR32(re_cache, mb->partial_bit_count); + re_index = last_index; + break; + } + re_index += vlc_len; #ifdef VLC_DEBUG - printf("run=%d level=%d\n", run, level); + printf("run=%d level=%d\n", run, level); #endif - pos += run; - if (pos >= 64) - break; + pos += run; + if (pos >= 64) + break; assert(level); pos1 = scan_table[pos]; @@ -404,7 +404,7 @@ block = block1; for(j = 0;j < 6; j++) { last_index = block_sizes[j]; - init_get_bits(&gb, buf_ptr, last_index); + init_get_bits(&gb, buf_ptr, last_index); /* get the dc */ dc = get_sbits(&gb, 9); @@ -444,7 +444,7 @@ block = block1; mb = mb1; init_get_bits(&gb, mb_bit_buffer, put_bits_count(&pb)); - flush_put_bits(&pb); + flush_put_bits(&pb); for(j = 0;j < 6; j++, block += 64, mb++) { if (mb->pos < 64 && get_bits_left(&gb) > 0) { dv_decode_ac(&gb, mb, block); @@ -456,7 +456,7 @@ /* all blocks are finished, so the extra bytes can be used at the video segment level */ if (j >= 6) - bit_copy(&vs_pb, &gb); + bit_copy(&vs_pb, &gb); } /* we need a pass other the whole video segment */ @@ -475,8 +475,8 @@ #endif dv_decode_ac(&gb, mb, block); } - if (mb->pos >= 64 && mb->pos < 127) - av_log(NULL, AV_LOG_ERROR, "AC EOB marker is absent pos=%d\n", mb->pos); + if (mb->pos >= 64 && mb->pos < 127) + av_log(NULL, AV_LOG_ERROR, "AC EOB marker is absent pos=%d\n", mb->pos); block += 64; mb++; } @@ -508,7 +508,7 @@ if (s->sys->pix_fmt == PIX_FMT_YUV411P && mb_x >= (704 / 8)) { uint64_t aligned_pixels[64/8]; uint8_t *pixels= (uint8_t*)aligned_pixels; - uint8_t *c_ptr, *c_ptr1, *ptr, *ptr1; + uint8_t *c_ptr, *c_ptr1, *ptr, *ptr1; int x, y, linesize; /* NOTE: at end of line, the macroblock is handled as 420 */ idct_put(pixels, 8, block); @@ -543,21 +543,21 @@ int size; if (run < DV_VLC_MAP_RUN_SIZE && level < DV_VLC_MAP_LEV_SIZE) { *vlc = dv_vlc_map[run][level].vlc | sign; - size = dv_vlc_map[run][level].size; + size = dv_vlc_map[run][level].size; } else { if (level < DV_VLC_MAP_LEV_SIZE) { - *vlc = dv_vlc_map[0][level].vlc | sign; - size = dv_vlc_map[0][level].size; - } else { + *vlc = dv_vlc_map[0][level].vlc | sign; + size = dv_vlc_map[0][level].size; + } else { *vlc = 0xfe00 | (level << 1) | sign; - size = 16; - } - if (run) { - *vlc |= ((run < 16) ? dv_vlc_map[run-1][0].vlc : - (0x1f80 | (run - 1))) << size; - size += (run < 16) ? dv_vlc_map[run-1][0].size : 13; - } + size = 16; + } + if (run) { + *vlc |= ((run < 16) ? dv_vlc_map[run-1][0].vlc : + (0x1f80 | (run - 1))) << size; + size += (run < 16) ? dv_vlc_map[run-1][0].size : 13; + } } return size; @@ -568,13 +568,13 @@ int size; if (run < DV_VLC_MAP_RUN_SIZE && level < DV_VLC_MAP_LEV_SIZE) { - size = dv_vlc_map[run][level].size; + size = dv_vlc_map[run][level].size; } else { - size = (level < DV_VLC_MAP_LEV_SIZE) ? dv_vlc_map[0][level].size : 16; - if (run) { - size += (run < 16) ? dv_vlc_map[run-1][0].size : 13; - } + size = (level < DV_VLC_MAP_LEV_SIZE) ? dv_vlc_map[0][level].size : 16; + if (run) { + size += (run < 16) ? dv_vlc_map[run-1][0].size : 13; + } } return size; } @@ -620,14 +620,14 @@ for (; size > (bits_left = put_bits_left(pb)); pb++) { if (bits_left) { size -= bits_left; - put_bits(pb, bits_left, vlc >> size); - vlc = vlc & ((1<<size)-1); - } - if (pb + 1 >= pb_end) { - bi->partial_bit_count = size; - bi->partial_bit_buffer = vlc; - return pb; - } + put_bits(pb, bits_left, vlc >> size); + vlc = vlc & ((1<<size)-1); + } + if (pb + 1 >= pb_end) { + bi->partial_bit_count = size; + bi->partial_bit_buffer = vlc; + return pb; + } } /* Store VLC */ @@ -712,14 +712,14 @@ s = blk; for(i=0; i<7; i++) { score88 += SC(0, 8) + SC(1, 9) + SC(2, 10) + SC(3, 11) + - SC(4, 12) + SC(5,13) + SC(6, 14) + SC(7, 15); + SC(4, 12) + SC(5,13) + SC(6, 14) + SC(7, 15); s += 8; } /* Compute 2-4-8 score (small values give a better chance for 2-4-8 DCT) */ s = blk; for(i=0; i<6; i++) { score248 += SC(0, 16) + SC(1,17) + SC(2, 18) + SC(3, 19) + - SC(4, 20) + SC(5,21) + SC(6, 22) + SC(7, 23); + SC(4, 20) + SC(5,21) + SC(6, 22) + SC(7, 23); s += 8; } @@ -736,30 +736,30 @@ b = blks; for (i=0; i<5; i++) { if (!qnos[i]) - continue; + continue; - qnos[i]--; - size[i] = 0; + qnos[i]--; + size[i] = 0; for (j=0; j<6; j++, b++) { - for (a=0; a<4; a++) { - if (b->area_q[a] != dv_quant_shifts[qnos[i] + dv_quant_offset[b->cno]][a]) { - b->bit_size[a] = 1; // 4 areas 4 bits for EOB :) - b->area_q[a]++; + for (a=0; a<4; a++) { + if (b->area_q[a] != dv_quant_shifts[qnos[i] + dv_quant_offset[b->cno]][a]) { + b->bit_size[a] = 1; // 4 areas 4 bits for EOB :) + b->area_q[a]++; prev= b->prev[a]; for (k= b->next[prev] ; k<mb_area_start[a+1]; k= b->next[k]) { - b->mb[k] >>= 1; - if (b->mb[k]) { + b->mb[k] >>= 1; + if (b->mb[k]) { b->bit_size[a] += dv_rl2vlc_size(k - prev - 1, b->mb[k]); - prev= k; + prev= k; } else { b->next[prev] = b->next[k]; } - } + } b->prev[a+1]= prev; - } - size[i] += b->bit_size[a]; - } - } + } + size[i] += b->bit_size[a]; + } + } } } while ((vs_total_ac_bits < size[0] + size[1] + size[2] + size[3] + size[4]) && (qnos[0]|qnos[1]|qnos[2]|qnos[3]|qnos[4])); @@ -797,68 +797,68 @@ mb_x = v & 0xff; mb_y = v >> 8; y_ptr = s->picture.data[0] + (mb_y * s->picture.linesize[0] * 8) + (mb_x * 8); - c_offset = (s->sys->pix_fmt == PIX_FMT_YUV411P) ? - ((mb_y * s->picture.linesize[1] * 8) + ((mb_x >> 2) * 8)) : - (((mb_y >> 1) * s->picture.linesize[1] * 8) + ((mb_x >> 1) * 8)); - do_edge_wrap = 0; - qnos[mb_index] = 15; /* No quantization */ + c_offset = (s->sys->pix_fmt == PIX_FMT_YUV411P) ? + ((mb_y * s->picture.linesize[1] * 8) + ((mb_x >> 2) * 8)) : + (((mb_y >> 1) * s->picture.linesize[1] * 8) + ((mb_x >> 1) * 8)); + do_edge_wrap = 0; + qnos[mb_index] = 15; /* No quantization */ ptr = dif + mb_index*80 + 4; for(j = 0;j < 6; j++) { if (j < 4) { /* Four Y blocks */ - /* NOTE: at end of line, the macroblock is handled as 420 */ - if (s->sys->pix_fmt == PIX_FMT_YUV411P && mb_x < (704 / 8)) { + /* NOTE: at end of line, the macroblock is handled as 420 */ + if (s->sys->pix_fmt == PIX_FMT_YUV411P && mb_x < (704 / 8)) { data = y_ptr + (j * 8); } else { data = y_ptr + ((j & 1) * 8) + ((j >> 1) * 8 * s->picture.linesize[0]); } - linesize = s->picture.linesize[0]; + linesize = s->picture.linesize[0]; } else { /* Cr and Cb blocks */ - /* don't ask Fabrice why they inverted Cb and Cr ! */ - data = s->picture.data[6 - j] + c_offset; - linesize = s->picture.linesize[6 - j]; - if (s->sys->pix_fmt == PIX_FMT_YUV411P && mb_x >= (704 / 8)) - do_edge_wrap = 1; - } + /* don't ask Fabrice why they inverted Cb and Cr ! */ + data = s->picture.data[6 - j] + c_offset; + linesize = s->picture.linesize[6 - j]; + if (s->sys->pix_fmt == PIX_FMT_YUV411P && mb_x >= (704 / 8)) + do_edge_wrap = 1; + } - /* Everything is set up -- now just copy data -> DCT block */ - if (do_edge_wrap) { /* Edge wrap copy: 4x16 -> 8x8 */ - uint8_t* d; - DCTELEM *b = block; - for (i=0;i<8;i++) { - d = data + 8 * linesize; - b[0] = data[0]; b[1] = data[1]; b[2] = data[2]; b[3] = data[3]; + /* Everything is set up -- now just copy data -> DCT block */ + if (do_edge_wrap) { /* Edge wrap copy: 4x16 -> 8x8 */ + uint8_t* d; + DCTELEM *b = block; + for (i=0;i<8;i++) { + d = data + 8 * linesize; + b[0] = data[0]; b[1] = data[1]; b[2] = data[2]; b[3] = data[3]; b[4] = d[0]; b[5] = d[1]; b[6] = d[2]; b[7] = d[3]; - data += linesize; - b += 8; - } - } else { /* Simple copy: 8x8 -> 8x8 */ - s->get_pixels(block, data, linesize); - } + data += linesize; + b += 8; + } + } else { /* Simple copy: 8x8 -> 8x8 */ + s->get_pixels(block, data, linesize); + } if(s->avctx->flags & CODEC_FLAG_INTERLACED_DCT) enc_blk->dct_mode = dv_guess_dct_mode(block); else enc_blk->dct_mode = 0; - enc_blk->area_q[0] = enc_blk->area_q[1] = enc_blk->area_q[2] = enc_blk->area_q[3] = 0; - enc_blk->partial_bit_count = 0; - enc_blk->partial_bit_buffer = 0; - enc_blk->cur_ac = 0; + enc_blk->area_q[0] = enc_blk->area_q[1] = enc_blk->area_q[2] = enc_blk->area_q[3] = 0; + enc_blk->partial_bit_count = 0; + enc_blk->partial_bit_buffer = 0; + enc_blk->cur_ac = 0; - s->fdct[enc_blk->dct_mode](block); + s->fdct[enc_blk->dct_mode](block); - dv_set_class_number(block, enc_blk, - enc_blk->dct_mode ? ff_zigzag248_direct : ff_zigzag_direct, j/4); + dv_set_class_number(block, enc_blk, + enc_blk->dct_mode ? ff_zigzag248_direct : ff_zigzag_direct, j/4); init_put_bits(pb, ptr, block_sizes[j]/8); - put_bits(pb, 9, (uint16_t)(((enc_blk->mb[0] >> 3) - 1024 + 2) >> 2)); - put_bits(pb, 1, enc_blk->dct_mode); - put_bits(pb, 2, enc_blk->cno); + put_bits(pb, 9, (uint16_t)(((enc_blk->mb[0] >> 3) - 1024 + 2) >> 2)); + put_bits(pb, 1, enc_blk->dct_mode); + put_bits(pb, 2, enc_blk->cno); - vs_bit_size += enc_blk->bit_size[0] + enc_blk->bit_size[1] + - enc_blk->bit_size[2] + enc_blk->bit_size[3]; - ++enc_blk; - ++pb; - ptr += block_sizes[j]/8; + vs_bit_size += enc_blk->bit_size[0] + enc_blk->bit_size[1] + + enc_blk->bit_size[2] + enc_blk->bit_size[3]; + ++enc_blk; + ++pb; + ptr += block_sizes[j]/8; } } @@ -898,7 +898,7 @@ DVVideoContext *s = avctx->priv_data; int slice = (size_t)sl; dv_decode_video_segment(s, &s->buf[((slice/27)*6+(slice/3)+slice*5+7)*80], - &s->sys->video_place[slice*5]); + &s->sys->video_place[slice*5]); return 0; } @@ -907,7 +907,7 @@ DVVideoContext *s = avctx->priv_data; int slice = (size_t)sl; dv_encode_video_segment(s, &s->buf[((slice/27)*6+(slice/3)+slice*5+7)*80], - &s->sys->video_place[slice*5]); + &s->sys->video_place[slice*5]); return 0; } @@ -940,7 +940,7 @@ s->buf = buf; avctx->execute(avctx, dv_decode_mt, (void**)&dv_anchor[0], NULL, - s->sys->difseg_size * 27); + s->sys->difseg_size * 27); emms_c(); @@ -958,7 +958,7 @@ s->sys = dv_codec_profile(c); if (!s->sys) - return -1; + return -1; if(buf_size < s->sys->frame_size) return -1; @@ -969,7 +969,7 @@ s->buf = buf; c->execute(c, dv_encode_mt, (void**)&dv_anchor[0], NULL, - s->sys->difseg_size * 27); + s->sys->difseg_size * 27); emms_c(); return s->sys->frame_size;
--- a/dvbsub.c Wed Dec 21 17:50:40 2005 +0000 +++ b/dvbsub.c Thu Dec 22 01:10:11 2005 +0000 @@ -192,7 +192,7 @@ #define SCALEBITS 10 #define ONE_HALF (1 << (SCALEBITS - 1)) -#define FIX(x) ((int) ((x) * (1<<SCALEBITS) + 0.5)) +#define FIX(x) ((int) ((x) * (1<<SCALEBITS) + 0.5)) #define RGB_TO_Y_CCIR(r, g, b) \ ((FIX(0.29900*219.0/255.0) * (r) + FIX(0.58700*219.0/255.0) * (g) + \
--- a/error_resilience.c Wed Dec 21 17:50:40 2005 +0000 +++ b/error_resilience.c Thu Dec 22 01:10:11 2005 +0000 @@ -108,8 +108,8 @@ /** * guess the dc of blocks which dont have a undamaged dc - * @param w width in 8 pixel blocks - * @param h height in 8 pixel blocks + * @param w width in 8 pixel blocks + * @param h height in 8 pixel blocks */ static void guess_dc(MpegEncContext *s, int16_t *dc, int w, int h, int stride, int is_luma){ int b_x, b_y; @@ -192,8 +192,8 @@ /** * simple horizontal deblocking filter used for error resilience - * @param w width in 8 pixel blocks - * @param h height in 8 pixel blocks + * @param w width in 8 pixel blocks + * @param h height in 8 pixel blocks */ static void h_block_filter(MpegEncContext *s, uint8_t *dst, int w, int h, int stride, int is_luma){ int b_x, b_y; @@ -252,8 +252,8 @@ /** * simple vertical deblocking filter used for error resilience - * @param w width in 8 pixel blocks - * @param h height in 8 pixel blocks + * @param w width in 8 pixel blocks + * @param h height in 8 pixel blocks */ static void v_block_filter(MpegEncContext *s, uint8_t *dst, int w, int h, int stride, int is_luma){ int b_x, b_y; @@ -348,7 +348,7 @@ s->mv_type = MV_TYPE_16X16; s->mb_skipped=0; - s->dsp.clear_blocks(s->block[0]); + s->dsp.clear_blocks(s->block[0]); s->mb_x= mb_x; s->mb_y= mb_y; @@ -476,7 +476,7 @@ s->mv_type = MV_TYPE_16X16; s->mb_skipped=0; - s->dsp.clear_blocks(s->block[0]); + s->dsp.clear_blocks(s->block[0]); s->mb_x= mb_x; s->mb_y= mb_y; @@ -582,7 +582,7 @@ uint8_t *mb_ptr = s->current_picture.data[0] + mb_x*16 + mb_y*16*s->linesize; uint8_t *last_mb_ptr= s->last_picture.data [0] + mb_x*16 + mb_y*16*s->linesize; - is_intra_likely += s->dsp.sad[0](NULL, last_mb_ptr, mb_ptr , s->linesize, 16); + is_intra_likely += s->dsp.sad[0](NULL, last_mb_ptr, mb_ptr , s->linesize, 16); is_intra_likely -= s->dsp.sad[0](NULL, last_mb_ptr, last_mb_ptr+s->linesize*16, s->linesize, 16); }else{ if(IS_INTRA(s->current_picture.mb_type[mb_xy])) @@ -873,7 +873,7 @@ s->mv[0][0][1] = s->current_picture.motion_val[0][ mb_x*2 + mb_y*2*s->b8_stride ][1]; } - s->dsp.clear_blocks(s->block[0]); + s->dsp.clear_blocks(s->block[0]); s->mb_x= mb_x; s->mb_y= mb_y;
--- a/faac.c Wed Dec 21 17:50:40 2005 +0000 +++ b/faac.c Thu Dec 22 01:10:11 2005 +0000 @@ -46,7 +46,7 @@ /* check faac version */ faac_cfg = faacEncGetCurrentConfiguration(s->faac_handle); if (faac_cfg->version != FAAC_CFG_VERSION) { - av_log(avctx, AV_LOG_ERROR, "wrong libfaac version (compiled for: %d, using %d)\n", FAAC_CFG_VERSION, faac_cfg->version); + av_log(avctx, AV_LOG_ERROR, "wrong libfaac version (compiled for: %d, using %d)\n", FAAC_CFG_VERSION, faac_cfg->version); faacEncClose(s->faac_handle); return -1; }
--- a/faad.c Wed Dec 21 17:50:40 2005 +0000 +++ b/faad.c Thu Dec 22 01:10:11 2005 +0000 @@ -47,8 +47,8 @@ #endif typedef struct { - void* handle; /* dlopen handle */ - void* faac_handle; /* FAAD library handle */ + void* handle; /* dlopen handle */ + void* faac_handle; /* FAAD library handle */ int frame_size; int sample_size; int flags; @@ -57,36 +57,36 @@ faacDecHandle FAADAPI (*faacDecOpen)(void); faacDecConfigurationPtr FAADAPI (*faacDecGetCurrentConfiguration)(faacDecHandle hDecoder); #ifndef FAAD2_VERSION - int FAADAPI (*faacDecSetConfiguration)(faacDecHandle hDecoder, + int FAADAPI (*faacDecSetConfiguration)(faacDecHandle hDecoder, faacDecConfigurationPtr config); - int FAADAPI (*faacDecInit)(faacDecHandle hDecoder, - unsigned char *buffer, - unsigned long *samplerate, - unsigned long *channels); - int FAADAPI (*faacDecInit2)(faacDecHandle hDecoder, unsigned char *pBuffer, + int FAADAPI (*faacDecInit)(faacDecHandle hDecoder, + unsigned char *buffer, + unsigned long *samplerate, + unsigned long *channels); + int FAADAPI (*faacDecInit2)(faacDecHandle hDecoder, unsigned char *pBuffer, unsigned long SizeOfDecoderSpecificInfo, unsigned long *samplerate, unsigned long *channels); - int FAADAPI (*faacDecDecode)(faacDecHandle hDecoder, - unsigned char *buffer, - unsigned long *bytesconsumed, - short *sample_buffer, + int FAADAPI (*faacDecDecode)(faacDecHandle hDecoder, + unsigned char *buffer, + unsigned long *bytesconsumed, + short *sample_buffer, unsigned long *samples); #else - unsigned char FAADAPI (*faacDecSetConfiguration)(faacDecHandle hDecoder, + unsigned char FAADAPI (*faacDecSetConfiguration)(faacDecHandle hDecoder, faacDecConfigurationPtr config); - long FAADAPI (*faacDecInit)(faacDecHandle hDecoder, - unsigned char *buffer, - unsigned long buffer_size, - unsigned long *samplerate, - unsigned char *channels); - char FAADAPI (*faacDecInit2)(faacDecHandle hDecoder, unsigned char *pBuffer, + long FAADAPI (*faacDecInit)(faacDecHandle hDecoder, + unsigned char *buffer, + unsigned long buffer_size, + unsigned long *samplerate, + unsigned char *channels); + char FAADAPI (*faacDecInit2)(faacDecHandle hDecoder, unsigned char *pBuffer, unsigned long SizeOfDecoderSpecificInfo, unsigned long *samplerate, unsigned char *channels); - void *FAADAPI (*faacDecDecode)(faacDecHandle hDecoder, - faacDecFrameInfo *hInfo, - unsigned char *buffer, - unsigned long buffer_size); - char* FAADAPI (*faacDecGetErrorMessage)(unsigned char errcode); + void *FAADAPI (*faacDecDecode)(faacDecHandle hDecoder, + faacDecFrameInfo *hInfo, + unsigned char *buffer, + unsigned long buffer_size); + char* FAADAPI (*faacDecGetErrorMessage)(unsigned char errcode); #endif void FAADAPI (*faacDecClose)(faacDecHandle hDecoder); @@ -112,14 +112,14 @@ int r = 0; if (avctx->extradata) - r = s->faacDecInit2(s->faac_handle, (uint8_t*) avctx->extradata, - avctx->extradata_size, - &samplerate, &channels); + r = s->faacDecInit2(s->faac_handle, (uint8_t*) avctx->extradata, + avctx->extradata_size, + &samplerate, &channels); // else r = s->faacDecInit(s->faac_handle ... ); if (r < 0) - av_log(avctx, AV_LOG_ERROR, "faacDecInit2 failed r:%d sr:%ld ch:%ld s:%d\n", - r, samplerate, (long)channels, avctx->extradata_size); + av_log(avctx, AV_LOG_ERROR, "faacDecInit2 failed r:%d sr:%ld ch:%ld s:%d\n", + r, samplerate, (long)channels, avctx->extradata_size); avctx->sample_rate = samplerate; avctx->channels = channels; @@ -141,7 +141,7 @@ void *out; #endif if(buf_size == 0) - return 0; + return 0; #ifndef FAAD2_VERSION out = s->faacDecDecode(s->faac_handle, (unsigned char*)buf, @@ -150,16 +150,16 @@ &samples); samples *= s->sample_size; if (data_size) - *data_size = samples; + *data_size = samples; return (buf_size < (int)bytesconsumed) - ? buf_size : (int)bytesconsumed; + ? buf_size : (int)bytesconsumed; #else out = s->faacDecDecode(s->faac_handle, &frame_info, (unsigned char*)buf, (unsigned long)buf_size); if (frame_info.error > 0) { - av_log(avctx, AV_LOG_ERROR, "faac: frame decoding failed: %s\n", - s->faacDecGetErrorMessage(frame_info.error)); + av_log(avctx, AV_LOG_ERROR, "faac: frame decoding failed: %s\n", + s->faacDecGetErrorMessage(frame_info.error)); return 0; } @@ -167,10 +167,10 @@ memcpy(data, out, frame_info.samples); // CHECKME - can we cheat this one if (data_size) - *data_size = frame_info.samples; + *data_size = frame_info.samples; return (buf_size < (int)frame_info.bytesconsumed) - ? buf_size : (int)frame_info.bytesconsumed; + ? buf_size : (int)frame_info.bytesconsumed; #endif } @@ -196,8 +196,8 @@ s->handle = dlopen(libfaadname, RTLD_LAZY); if (!s->handle) { - av_log(avctx, AV_LOG_ERROR, "FAAD library: %s could not be opened! \n%s\n", - libfaadname, dlerror()); + av_log(avctx, AV_LOG_ERROR, "FAAD library: %s could not be opened! \n%s\n", + libfaadname, dlerror()); return -1; } #define dfaac(a, b) \ @@ -209,32 +209,32 @@ #endif /* CONFIG_FAADBIN */ // resolve all needed function calls - dfaac(Open, (faacDecHandle FAADAPI (*)(void))); - dfaac(GetCurrentConfiguration, (faacDecConfigurationPtr - FAADAPI (*)(faacDecHandle))); + dfaac(Open, (faacDecHandle FAADAPI (*)(void))); + dfaac(GetCurrentConfiguration, (faacDecConfigurationPtr + FAADAPI (*)(faacDecHandle))); #ifndef FAAD2_VERSION - dfaac(SetConfiguration, (int FAADAPI (*)(faacDecHandle, - faacDecConfigurationPtr))); + dfaac(SetConfiguration, (int FAADAPI (*)(faacDecHandle, + faacDecConfigurationPtr))); - dfaac(Init, (int FAADAPI (*)(faacDecHandle, unsigned char*, - unsigned long*, unsigned long*))); + dfaac(Init, (int FAADAPI (*)(faacDecHandle, unsigned char*, + unsigned long*, unsigned long*))); dfaac(Init2, (int FAADAPI (*)(faacDecHandle, unsigned char*, - unsigned long, unsigned long*, - unsigned long*))); + unsigned long, unsigned long*, + unsigned long*))); dfaac(Close, (void FAADAPI (*)(faacDecHandle hDecoder))); - dfaac(Decode, (int FAADAPI (*)(faacDecHandle, unsigned char*, - unsigned long*, short*, unsigned long*))); + dfaac(Decode, (int FAADAPI (*)(faacDecHandle, unsigned char*, + unsigned long*, short*, unsigned long*))); #else - dfaac(SetConfiguration, (unsigned char FAADAPI (*)(faacDecHandle, - faacDecConfigurationPtr))); - dfaac(Init, (long FAADAPI (*)(faacDecHandle, unsigned char*, - unsigned long, unsigned long*, unsigned char*))); - dfaac(Init2, (char FAADAPI (*)(faacDecHandle, unsigned char*, - unsigned long, unsigned long*, - unsigned char*))); - dfaac(Decode, (void *FAADAPI (*)(faacDecHandle, faacDecFrameInfo*, - unsigned char*, unsigned long))); - dfaac(GetErrorMessage, (char* FAADAPI (*)(unsigned char))); + dfaac(SetConfiguration, (unsigned char FAADAPI (*)(faacDecHandle, + faacDecConfigurationPtr))); + dfaac(Init, (long FAADAPI (*)(faacDecHandle, unsigned char*, + unsigned long, unsigned long*, unsigned char*))); + dfaac(Init2, (char FAADAPI (*)(faacDecHandle, unsigned char*, + unsigned long, unsigned long*, + unsigned char*))); + dfaac(Decode, (void *FAADAPI (*)(faacDecHandle, faacDecFrameInfo*, + unsigned char*, unsigned long))); + dfaac(GetErrorMessage, (char* FAADAPI (*)(unsigned char))); #endif #undef dfacc @@ -243,8 +243,8 @@ } if (err) { dlclose(s->handle); - av_log(avctx, AV_LOG_ERROR, "FAAD library: cannot resolve %s in %s!\n", - err, libfaadname); + av_log(avctx, AV_LOG_ERROR, "FAAD library: cannot resolve %s in %s!\n", + err, libfaadname); return -1; } #endif @@ -260,31 +260,31 @@ faac_cfg = s->faacDecGetCurrentConfiguration(s->faac_handle); if (faac_cfg) { - switch (avctx->bits_per_sample) { - case 8: av_log(avctx, AV_LOG_ERROR, "FAADlib unsupported bps %d\n", avctx->bits_per_sample); break; - default: - case 16: + switch (avctx->bits_per_sample) { + case 8: av_log(avctx, AV_LOG_ERROR, "FAADlib unsupported bps %d\n", avctx->bits_per_sample); break; + default: + case 16: #ifdef FAAD2_VERSION - faac_cfg->outputFormat = FAAD_FMT_16BIT; + faac_cfg->outputFormat = FAAD_FMT_16BIT; #endif - s->sample_size = 2; - break; - case 24: + s->sample_size = 2; + break; + case 24: #ifdef FAAD2_VERSION - faac_cfg->outputFormat = FAAD_FMT_24BIT; + faac_cfg->outputFormat = FAAD_FMT_24BIT; #endif - s->sample_size = 3; - break; - case 32: + s->sample_size = 3; + break; + case 32: #ifdef FAAD2_VERSION - faac_cfg->outputFormat = FAAD_FMT_32BIT; + faac_cfg->outputFormat = FAAD_FMT_32BIT; #endif - s->sample_size = 4; - break; - } + s->sample_size = 4; + break; + } - faac_cfg->defSampleRate = (!avctx->sample_rate) ? 44100 : avctx->sample_rate; - faac_cfg->defObjectType = LC; + faac_cfg->defSampleRate = (!avctx->sample_rate) ? 44100 : avctx->sample_rate; + faac_cfg->defObjectType = LC; } s->faacDecSetConfiguration(s->faac_handle, faac_cfg);
--- a/faandct.c Wed Dec 21 17:50:40 2005 +0000 +++ b/faandct.c Thu Dec 22 01:10:11 2005 +0000 @@ -204,15 +204,15 @@ data[8*6 + i] = lrintf(SCALE(8*6 + i) * (tmp13 - z1)); tmp10 = tmp4 + tmp7; - tmp11 = tmp5 + tmp6; - tmp12 = tmp5 - tmp6; - tmp13 = tmp4 - tmp7; + tmp11 = tmp5 + tmp6; + tmp12 = tmp5 - tmp6; + tmp13 = tmp4 - tmp7; - data[8*1 + i] = lrintf(SCALE(8*0 + i) * (tmp10 + tmp11)); - data[8*5 + i] = lrintf(SCALE(8*4 + i) * (tmp10 - tmp11)); + data[8*1 + i] = lrintf(SCALE(8*0 + i) * (tmp10 + tmp11)); + data[8*5 + i] = lrintf(SCALE(8*4 + i) * (tmp10 - tmp11)); - z1 = (tmp12 + tmp13)* A1; - data[8*3 + i] = lrintf(SCALE(8*2 + i) * (tmp13 + z1)); - data[8*7 + i] = lrintf(SCALE(8*6 + i) * (tmp13 - z1)); + z1 = (tmp12 + tmp13)* A1; + data[8*3 + i] = lrintf(SCALE(8*2 + i) * (tmp13 + z1)); + data[8*7 + i] = lrintf(SCALE(8*6 + i) * (tmp13 - z1)); } }
--- a/fdctref.c Wed Dec 21 17:50:40 2005 +0000 +++ b/fdctref.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,51 +64,51 @@ void fdct(block) short *block; { - register int i, j; - double s; - double tmp[64]; + register int i, j; + double s; + double tmp[64]; - for(i = 0; i < 8; i++) - for(j = 0; j < 8; j++) - { - s = 0.0; + for(i = 0; i < 8; i++) + for(j = 0; j < 8; j++) + { + s = 0.0; /* - * for(k = 0; k < 8; k++) - * s += c[j][k] * block[8 * i + k]; + * for(k = 0; k < 8; k++) + * s += c[j][k] * block[8 * i + k]; */ - s += c[j][0] * block[8 * i + 0]; - s += c[j][1] * block[8 * i + 1]; - s += c[j][2] * block[8 * i + 2]; - s += c[j][3] * block[8 * i + 3]; - s += c[j][4] * block[8 * i + 4]; - s += c[j][5] * block[8 * i + 5]; - s += c[j][6] * block[8 * i + 6]; - s += c[j][7] * block[8 * i + 7]; + s += c[j][0] * block[8 * i + 0]; + s += c[j][1] * block[8 * i + 1]; + s += c[j][2] * block[8 * i + 2]; + s += c[j][3] * block[8 * i + 3]; + s += c[j][4] * block[8 * i + 4]; + s += c[j][5] * block[8 * i + 5]; + s += c[j][6] * block[8 * i + 6]; + s += c[j][7] * block[8 * i + 7]; - tmp[8 * i + j] = s; - } + tmp[8 * i + j] = s; + } - for(j = 0; j < 8; j++) - for(i = 0; i < 8; i++) - { - s = 0.0; + for(j = 0; j < 8; j++) + for(i = 0; i < 8; i++) + { + s = 0.0; /* - * for(k = 0; k < 8; k++) - * s += c[i][k] * tmp[8 * k + j]; + * for(k = 0; k < 8; k++) + * s += c[i][k] * tmp[8 * k + j]; */ - s += c[i][0] * tmp[8 * 0 + j]; - s += c[i][1] * tmp[8 * 1 + j]; - s += c[i][2] * tmp[8 * 2 + j]; - s += c[i][3] * tmp[8 * 3 + j]; - s += c[i][4] * tmp[8 * 4 + j]; - s += c[i][5] * tmp[8 * 5 + j]; - s += c[i][6] * tmp[8 * 6 + j]; - s += c[i][7] * tmp[8 * 7 + j]; - s*=8.0; + s += c[i][0] * tmp[8 * 0 + j]; + s += c[i][1] * tmp[8 * 1 + j]; + s += c[i][2] * tmp[8 * 2 + j]; + s += c[i][3] * tmp[8 * 3 + j]; + s += c[i][4] * tmp[8 * 4 + j]; + s += c[i][5] * tmp[8 * 5 + j]; + s += c[i][6] * tmp[8 * 6 + j]; + s += c[i][7] * tmp[8 * 7 + j]; + s*=8.0; - block[8 * i + j] = (short)floor(s + 0.499999); + block[8 * i + j] = (short)floor(s + 0.499999); /* * reason for adding 0.499999 instead of 0.5: * s is quite often x.5 (at least for i and/or j = 0 or 4)
--- a/fft.c Wed Dec 21 17:50:40 2005 +0000 +++ b/fft.c Thu Dec 22 01:10:11 2005 +0000 @@ -149,8 +149,8 @@ void ff_fft_calc_c(FFTContext *s, FFTComplex *z) { int ln = s->nbits; - int j, np, np2; - int nblocks, nloops; + int j, np, np2; + int nblocks, nloops; register FFTComplex *p, *q; FFTComplex *exptab = s->exptab; int l;
--- a/g726.c Wed Dec 21 17:50:40 2005 +0000 +++ b/g726.c Thu Dec 22 01:10:11 2005 +0000 @@ -31,30 +31,30 @@ * instead of simply using 32bit integer arithmetic. */ typedef struct Float11 { - int sign; /**< 1bit sign */ - int exp; /**< 4bit exponent */ - int mant; /**< 6bit mantissa */ + int sign; /**< 1bit sign */ + int exp; /**< 4bit exponent */ + int mant; /**< 6bit mantissa */ } Float11; static inline Float11* i2f(int16_t i, Float11* f) { - f->sign = (i < 0); - if (f->sign) - i = -i; - f->exp = av_log2_16bit(i) + !!i; - f->mant = i? (i<<6) >> f->exp : - 1<<5; - return f; + f->sign = (i < 0); + if (f->sign) + i = -i; + f->exp = av_log2_16bit(i) + !!i; + f->mant = i? (i<<6) >> f->exp : + 1<<5; + return f; } static inline int16_t mult(Float11* f1, Float11* f2) { - int res, exp; + int res, exp; - exp = f1->exp + f2->exp; - res = (((f1->mant * f2->mant) + 0x30) >> 4) << 7; - res = exp > 26 ? res << (exp - 26) : res >> (26 - exp); - return (f1->sign ^ f2->sign) ? -res : res; + exp = f1->exp + f2->exp; + res = (((f1->mant * f2->mant) + 0x30) >> 4) << 7; + res = exp > 26 ? res << (exp - 26) : res >> (26 - exp); + return (f1->sign ^ f2->sign) ? -res : res; } static inline int sgn(int value) @@ -63,32 +63,32 @@ } typedef struct G726Tables { - int bits; /**< bits per sample */ - int* quant; /**< quantization table */ - int* iquant; /**< inverse quantization table */ - int* W; /**< special table #1 ;-) */ - int* F; /**< special table #2 */ + int bits; /**< bits per sample */ + int* quant; /**< quantization table */ + int* iquant; /**< inverse quantization table */ + int* W; /**< special table #1 ;-) */ + int* F; /**< special table #2 */ } G726Tables; typedef struct G726Context { - G726Tables* tbls; /**< static tables needed for computation */ + G726Tables* tbls; /**< static tables needed for computation */ - Float11 sr[2]; /**< prev. reconstructed samples */ - Float11 dq[6]; /**< prev. difference */ - int a[2]; /**< second order predictor coeffs */ - int b[6]; /**< sixth order predictor coeffs */ - int pk[2]; /**< signs of prev. 2 sez + dq */ + Float11 sr[2]; /**< prev. reconstructed samples */ + Float11 dq[6]; /**< prev. difference */ + int a[2]; /**< second order predictor coeffs */ + int b[6]; /**< sixth order predictor coeffs */ + int pk[2]; /**< signs of prev. 2 sez + dq */ - int ap; /**< scale factor control */ - int yu; /**< fast scale factor */ - int yl; /**< slow scale factor */ - int dms; /**< short average magnitude of F[i] */ - int dml; /**< long average magnitude of F[i] */ - int td; /**< tone detect */ + int ap; /**< scale factor control */ + int yu; /**< fast scale factor */ + int yl; /**< slow scale factor */ + int dms; /**< short average magnitude of F[i] */ + int dml; /**< long average magnitude of F[i] */ + int td; /**< tone detect */ - int se; /**< estimated signal for the next iteration */ - int sez; /**< estimated second order prediction */ - int y; /**< quantizer scaling factor for the next iteration */ + int se; /**< estimated signal for the next iteration */ + int sez; /**< estimated second order prediction */ + int y; /**< quantizer scaling factor for the next iteration */ } G726Context; static int quant_tbl16[] = /**< 16kbit/s 2bits per sample */ @@ -113,34 +113,34 @@ { -125, 79, 177, 245, 299, 348, 399, INT_MAX }; static int iquant_tbl32[] = { INT_MIN, 4, 135, 213, 273, 323, 373, 425, - 425, 373, 323, 273, 213, 135, 4, INT_MIN }; + 425, 373, 323, 273, 213, 135, 4, INT_MIN }; static int W_tbl32[] = { -12, 18, 41, 64, 112, 198, 355, 1122, - 1122, 355, 198, 112, 64, 41, 18, -12}; + 1122, 355, 198, 112, 64, 41, 18, -12}; static int F_tbl32[] = { 0, 0, 0, 1, 1, 1, 3, 7, 7, 3, 1, 1, 1, 0, 0, 0 }; static int quant_tbl40[] = /**< 40kbit/s 5bits per sample */ { -122, -16, 67, 138, 197, 249, 297, 338, - 377, 412, 444, 474, 501, 527, 552, INT_MAX }; + 377, 412, 444, 474, 501, 527, 552, INT_MAX }; static int iquant_tbl40[] = { INT_MIN, -66, 28, 104, 169, 224, 274, 318, - 358, 395, 429, 459, 488, 514, 539, 566, - 566, 539, 514, 488, 459, 429, 395, 358, - 318, 274, 224, 169, 104, 28, -66, INT_MIN }; + 358, 395, 429, 459, 488, 514, 539, 566, + 566, 539, 514, 488, 459, 429, 395, 358, + 318, 274, 224, 169, 104, 28, -66, INT_MIN }; static int W_tbl40[] = { 14, 14, 24, 39, 40, 41, 58, 100, - 141, 179, 219, 280, 358, 440, 529, 696, - 696, 529, 440, 358, 280, 219, 179, 141, - 100, 58, 41, 40, 39, 24, 14, 14 }; + 141, 179, 219, 280, 358, 440, 529, 696, + 696, 529, 440, 358, 280, 219, 179, 141, + 100, 58, 41, 40, 39, 24, 14, 14 }; static int F_tbl40[] = { 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 2, 3, 4, 5, 6, 6, - 6, 6, 5, 4, 3, 2, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0 }; + 6, 6, 5, 4, 3, 2, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0 }; static G726Tables G726Tables_pool[] = {{ 2, quant_tbl16, iquant_tbl16, W_tbl16, F_tbl16 }, - { 3, quant_tbl24, iquant_tbl24, W_tbl24, F_tbl24 }, - { 4, quant_tbl32, iquant_tbl32, W_tbl32, F_tbl32 }, + { 3, quant_tbl24, iquant_tbl24, W_tbl24, F_tbl24 }, + { 4, quant_tbl32, iquant_tbl32, W_tbl32, F_tbl32 }, { 5, quant_tbl40, iquant_tbl40, W_tbl40, F_tbl40 }}; @@ -207,20 +207,20 @@ dq0 = dq ? sgn(dq) : 0; if (tr) { c->a[0] = 0; - c->a[1] = 0; + c->a[1] = 0; for (i=0; i<6; i++) - c->b[i] = 0; + c->b[i] = 0; } else { - /* This is a bit crazy, but it really is +255 not +256 */ - fa1 = clip((-c->a[0]*c->pk[0]*pk0)>>5, -256, 255); + /* This is a bit crazy, but it really is +255 not +256 */ + fa1 = clip((-c->a[0]*c->pk[0]*pk0)>>5, -256, 255); - c->a[1] += 128*pk0*c->pk[1] + fa1 - (c->a[1]>>7); - c->a[1] = clip(c->a[1], -12288, 12288); + c->a[1] += 128*pk0*c->pk[1] + fa1 - (c->a[1]>>7); + c->a[1] = clip(c->a[1], -12288, 12288); c->a[0] += 64*3*pk0*c->pk[0] - (c->a[0] >> 8); - c->a[0] = clip(c->a[0], -(15360 - c->a[1]), 15360 - c->a[1]); + c->a[0] = clip(c->a[0], -(15360 - c->a[1]), 15360 - c->a[1]); for (i=0; i<6; i++) - c->b[i] += 128*dq0*sgn(-c->dq[i].sign) - (c->b[i]>>8); + c->b[i] += 128*dq0*sgn(-c->dq[i].sign) - (c->b[i]>>8); } /* Update Dq and Sr and Pk */ @@ -323,13 +323,13 @@ if (avctx->channels != 1 || (avctx->bit_rate != 16000 && avctx->bit_rate != 24000 && - avctx->bit_rate != 32000 && avctx->bit_rate != 40000)) { + avctx->bit_rate != 32000 && avctx->bit_rate != 40000)) { av_log(avctx, AV_LOG_ERROR, "G726: unsupported audio format\n"); - return -1; + return -1; } if (avctx->sample_rate != 8000 && avctx->strict_std_compliance>FF_COMPLIANCE_INOFFICIAL) { av_log(avctx, AV_LOG_ERROR, "G726: unsupported audio format\n"); - return -1; + return -1; } g726_reset(&c->c, avctx->bit_rate); c->code_size = c->c.tbls->bits; @@ -384,12 +384,12 @@ init_get_bits(&gb, buf, buf_size * 8); if (c->bits_left) { int s = c->code_size - c->bits_left;; - code = (c->bit_buffer << s) | get_bits(&gb, s); - *samples++ = g726_decode(&c->c, code & mask); + code = (c->bit_buffer << s) | get_bits(&gb, s); + *samples++ = g726_decode(&c->c, code & mask); } while (get_bits_count(&gb) + c->code_size <= buf_size*8) - *samples++ = g726_decode(&c->c, get_bits(&gb, c->code_size) & mask); + *samples++ = g726_decode(&c->c, get_bits(&gb, c->code_size) & mask); c->bits_left = buf_size*8 - get_bits_count(&gb); c->bit_buffer = get_bits(&gb, c->bits_left);
--- a/golomb.h Wed Dec 21 17:50:40 2005 +0000 +++ b/golomb.h Thu Dec 22 01:10:11 2005 +0000 @@ -288,7 +288,7 @@ * read unsigned golomb rice code (shorten). */ static inline unsigned int get_ur_golomb_shorten(GetBitContext *gb, int k){ - return get_ur_golomb_jpegls(gb, k, INT_MAX, 0); + return get_ur_golomb_jpegls(gb, k, INT_MAX, 0); } /** @@ -395,7 +395,7 @@ */ static inline void set_se_golomb(PutBitContext *pb, int i){ // if (i>32767 || i<-32767) -// av_log(NULL,AV_LOG_ERROR,"value out of range %d\n", i); +// av_log(NULL,AV_LOG_ERROR,"value out of range %d\n", i); #if 0 if(i<=0) i= -2*i; else i= 2*i-1;
--- a/h263.c Wed Dec 21 17:50:40 2005 +0000 +++ b/h263.c Thu Dec 22 01:10:11 2005 +0000 @@ -231,11 +231,11 @@ (coded_frame_rate_base * (int64_t)s->avctx->time_base.den); put_bits(&s->pb, 8, temp_ref & 0xff); /* TemporalReference */ - put_bits(&s->pb, 1, 1); /* marker */ - put_bits(&s->pb, 1, 0); /* h263 id */ - put_bits(&s->pb, 1, 0); /* split screen off */ - put_bits(&s->pb, 1, 0); /* camera off */ - put_bits(&s->pb, 1, 0); /* freeze picture release off */ + put_bits(&s->pb, 1, 1); /* marker */ + put_bits(&s->pb, 1, 0); /* h263 id */ + put_bits(&s->pb, 1, 0); /* split screen off */ + put_bits(&s->pb, 1, 0); /* camera off */ + put_bits(&s->pb, 1, 0); /* freeze picture release off */ format = h263_get_picture_format(s->width, s->height); if (!s->h263_plus) { @@ -245,12 +245,12 @@ /* By now UMV IS DISABLED ON H.263v1, since the restrictions of H.263v1 UMV implies to check the predicted MV after calculation of the current MB to see if we're on the limits */ - put_bits(&s->pb, 1, 0); /* Unrestricted Motion Vector: off */ - put_bits(&s->pb, 1, 0); /* SAC: off */ - put_bits(&s->pb, 1, s->obmc); /* Advanced Prediction */ - put_bits(&s->pb, 1, 0); /* only I/P frames, no PB frame */ + put_bits(&s->pb, 1, 0); /* Unrestricted Motion Vector: off */ + put_bits(&s->pb, 1, 0); /* SAC: off */ + put_bits(&s->pb, 1, s->obmc); /* Advanced Prediction */ + put_bits(&s->pb, 1, 0); /* only I/P frames, no PB frame */ put_bits(&s->pb, 5, s->qscale); - put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */ + put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */ } else { int ufep=1; /* H.263v2 */ @@ -286,9 +286,9 @@ put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */ /* This should be here if PLUSPTYPE */ - put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */ - - if (format == 7) { + put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */ + + if (format == 7) { /* Custom Picture Format (CPFMT) */ aspect_to_info(s, s->avctx->sample_aspect_ratio); @@ -299,7 +299,7 @@ if (s->aspect_ratio_info == FF_ASPECT_EXTENDED){ put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.num); put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.den); - } + } } if(s->custom_pcf){ if(ufep){ @@ -320,7 +320,7 @@ put_bits(&s->pb, 5, s->qscale); } - put_bits(&s->pb, 1, 0); /* no PEI */ + put_bits(&s->pb, 1, 0); /* no PEI */ if(s->h263_slice_structured){ put_bits(&s->pb, 1, 1); @@ -823,8 +823,8 @@ } void mpeg4_encode_mb(MpegEncContext * s, - DCTELEM block[6][64], - int motion_x, int motion_y) + DCTELEM block[6][64], + int motion_x, int motion_y) { int cbpc, cbpy, pred_x, pred_y; PutBitContext * const pb2 = s->data_partitioning ? &s->pb2 : &s->pb; @@ -884,7 +884,7 @@ return; } - put_bits(&s->pb, 1, 0); /* mb coded modb1=0 */ + put_bits(&s->pb, 1, 0); /* mb coded modb1=0 */ put_bits(&s->pb, 1, cbp ? 0 : 1); /* modb2 */ //FIXME merge put_bits(&s->pb, mb_type+1, 1); // this table is so simple that we don't need it :) if(cbp) put_bits(&s->pb, 6, cbp); @@ -998,7 +998,7 @@ if(pic==NULL || pic->pict_type!=B_TYPE) break; b_pic= pic->data[0] + offset + 16; //FIXME +16 - diff= s->dsp.sad[0](NULL, p_pic, b_pic, s->linesize, 16); + diff= s->dsp.sad[0](NULL, p_pic, b_pic, s->linesize, 16); if(diff>s->qscale*70){ //FIXME check that 70 is optimal s->mb_skipped=0; break; @@ -1021,7 +1021,7 @@ } } - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ cbpc = cbp & 3; cbpy = cbp >> 2; cbpy ^= 0xf; @@ -1121,7 +1121,7 @@ int dc_diff[6]; //dc values with the dc prediction subtracted int dir[6]; //prediction direction int zigzag_last_index[6]; - uint8_t *scan_table[6]; + uint8_t *scan_table[6]; int i; for(i=0; i<6; i++){ @@ -1152,7 +1152,7 @@ intra_MCBPC_code[cbpc]); } else { if(s->dquant) cbpc+=8; - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ put_bits(&s->pb, inter_MCBPC_bits[cbpc + 4], inter_MCBPC_code[cbpc + 4]); @@ -1185,8 +1185,8 @@ } void h263_encode_mb(MpegEncContext * s, - DCTELEM block[6][64], - int motion_x, int motion_y) + DCTELEM block[6][64], + int motion_x, int motion_y) { int cbpc, cbpy, i, cbp, pred_x, pred_y; int16_t pred_dc; @@ -1211,7 +1211,7 @@ return; } - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ cbpc = cbp & 3; cbpy = cbp >> 2; @@ -1346,14 +1346,14 @@ intra_MCBPC_code[cbpc]); } else { if(s->dquant) cbpc+=8; - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ put_bits(&s->pb, inter_MCBPC_bits[cbpc + 4], inter_MCBPC_code[cbpc + 4]); } if (s->h263_aic) { /* XXX: currently, we do not try to use ac prediction */ - put_bits(&s->pb, 1, 0); /* no AC prediction */ + put_bits(&s->pb, 1, 0); /* no AC prediction */ } cbpy = cbp >> 2; put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]); @@ -1796,7 +1796,7 @@ v = abs(level); while (v) { v >>= 1; - size++; + size++; } if (level < 0) @@ -2318,14 +2318,14 @@ put_bits(&s->pb, 16, 0); put_bits(&s->pb, 16, 0x120 + vol_number); /* video obj layer */ - put_bits(&s->pb, 1, 0); /* random access vol */ - put_bits(&s->pb, 8, s->vo_type); /* video obj type indication */ + put_bits(&s->pb, 1, 0); /* random access vol */ + put_bits(&s->pb, 8, s->vo_type); /* video obj type indication */ if(s->workaround_bugs & FF_BUG_MS) { - put_bits(&s->pb, 1, 0); /* is obj layer id= no */ + put_bits(&s->pb, 1, 0); /* is obj layer id= no */ } else { - put_bits(&s->pb, 1, 1); /* is obj layer id= yes */ - put_bits(&s->pb, 4, vo_ver_id); /* is obj layer ver id */ - put_bits(&s->pb, 3, 1); /* is obj layer priority */ + put_bits(&s->pb, 1, 1); /* is obj layer id= yes */ + put_bits(&s->pb, 4, vo_ver_id); /* is obj layer ver id */ + put_bits(&s->pb, 3, 1); /* is obj layer priority */ } aspect_to_info(s, s->avctx->sample_aspect_ratio); @@ -2337,37 +2337,37 @@ } if(s->workaround_bugs & FF_BUG_MS) { // - put_bits(&s->pb, 1, 0); /* vol control parameters= no @@@ */ + put_bits(&s->pb, 1, 0); /* vol control parameters= no @@@ */ } else { - put_bits(&s->pb, 1, 1); /* vol control parameters= yes */ - put_bits(&s->pb, 2, 1); /* chroma format YUV 420/YV12 */ + put_bits(&s->pb, 1, 1); /* vol control parameters= yes */ + put_bits(&s->pb, 2, 1); /* chroma format YUV 420/YV12 */ put_bits(&s->pb, 1, s->low_delay); - put_bits(&s->pb, 1, 0); /* vbv parameters= no */ - } - - put_bits(&s->pb, 2, RECT_SHAPE); /* vol shape= rectangle */ - put_bits(&s->pb, 1, 1); /* marker bit */ + put_bits(&s->pb, 1, 0); /* vbv parameters= no */ + } + + put_bits(&s->pb, 2, RECT_SHAPE); /* vol shape= rectangle */ + put_bits(&s->pb, 1, 1); /* marker bit */ put_bits(&s->pb, 16, s->avctx->time_base.den); if (s->time_increment_bits < 1) s->time_increment_bits = 1; - put_bits(&s->pb, 1, 1); /* marker bit */ - put_bits(&s->pb, 1, 0); /* fixed vop rate=no */ - put_bits(&s->pb, 1, 1); /* marker bit */ - put_bits(&s->pb, 13, s->width); /* vol width */ - put_bits(&s->pb, 1, 1); /* marker bit */ - put_bits(&s->pb, 13, s->height); /* vol height */ - put_bits(&s->pb, 1, 1); /* marker bit */ + put_bits(&s->pb, 1, 1); /* marker bit */ + put_bits(&s->pb, 1, 0); /* fixed vop rate=no */ + put_bits(&s->pb, 1, 1); /* marker bit */ + put_bits(&s->pb, 13, s->width); /* vol width */ + put_bits(&s->pb, 1, 1); /* marker bit */ + put_bits(&s->pb, 13, s->height); /* vol height */ + put_bits(&s->pb, 1, 1); /* marker bit */ put_bits(&s->pb, 1, s->progressive_sequence ? 0 : 1); - put_bits(&s->pb, 1, 1); /* obmc disable */ + put_bits(&s->pb, 1, 1); /* obmc disable */ if (vo_ver_id == 1) { - put_bits(&s->pb, 1, s->vol_sprite_usage); /* sprite enable */ + put_bits(&s->pb, 1, s->vol_sprite_usage); /* sprite enable */ }else{ - put_bits(&s->pb, 2, s->vol_sprite_usage); /* sprite enable */ - } - - put_bits(&s->pb, 1, 0); /* not 8 bit == false */ - put_bits(&s->pb, 1, s->mpeg_quant); /* quant type= (0=h263 style)*/ + put_bits(&s->pb, 2, s->vol_sprite_usage); /* sprite enable */ + } + + put_bits(&s->pb, 1, 0); /* not 8 bit == false */ + put_bits(&s->pb, 1, s->mpeg_quant); /* quant type= (0=h263 style)*/ if(s->mpeg_quant){ ff_write_quant_matrix(&s->pb, s->avctx->intra_matrix); @@ -2376,27 +2376,27 @@ if (vo_ver_id != 1) put_bits(&s->pb, 1, s->quarter_sample); - put_bits(&s->pb, 1, 1); /* complexity estimation disable */ + put_bits(&s->pb, 1, 1); /* complexity estimation disable */ s->resync_marker= s->rtp_mode; put_bits(&s->pb, 1, s->resync_marker ? 0 : 1);/* resync marker disable */ put_bits(&s->pb, 1, s->data_partitioning ? 1 : 0); if(s->data_partitioning){ - put_bits(&s->pb, 1, 0); /* no rvlc */ + put_bits(&s->pb, 1, 0); /* no rvlc */ } if (vo_ver_id != 1){ - put_bits(&s->pb, 1, 0); /* newpred */ - put_bits(&s->pb, 1, 0); /* reduced res vop */ - } - put_bits(&s->pb, 1, 0); /* scalability */ + put_bits(&s->pb, 1, 0); /* newpred */ + put_bits(&s->pb, 1, 0); /* reduced res vop */ + } + put_bits(&s->pb, 1, 0); /* scalability */ ff_mpeg4_stuffing(&s->pb); /* user data */ if(!(s->flags & CODEC_FLAG_BITEXACT)){ put_bits(&s->pb, 16, 0); - put_bits(&s->pb, 16, 0x1B2); /* user_data */ - ff_put_string(&s->pb, LIBAVCODEC_IDENT, 0); + put_bits(&s->pb, 16, 0x1B2); /* user_data */ + ff_put_string(&s->pb, LIBAVCODEC_IDENT, 0); } } @@ -2421,9 +2421,9 @@ //printf("num:%d rate:%d base:%d\n", s->picture_number, s->time_base.den, FRAME_RATE_BASE); - put_bits(&s->pb, 16, 0); /* vop header */ - put_bits(&s->pb, 16, VOP_STARTCODE); /* vop header */ - put_bits(&s->pb, 2, s->pict_type - 1); /* pict type: I = 0 , P = 1 */ + put_bits(&s->pb, 16, 0); /* vop header */ + put_bits(&s->pb, 16, VOP_STARTCODE); /* vop header */ + put_bits(&s->pb, 2, s->pict_type - 1); /* pict type: I = 0 , P = 1 */ assert(s->time>=0); time_div= s->time/s->avctx->time_base.den; @@ -2435,15 +2435,15 @@ put_bits(&s->pb, 1, 0); - put_bits(&s->pb, 1, 1); /* marker */ - put_bits(&s->pb, s->time_increment_bits, time_mod); /* time increment */ - put_bits(&s->pb, 1, 1); /* marker */ - put_bits(&s->pb, 1, 1); /* vop coded */ + put_bits(&s->pb, 1, 1); /* marker */ + put_bits(&s->pb, s->time_increment_bits, time_mod); /* time increment */ + put_bits(&s->pb, 1, 1); /* marker */ + put_bits(&s->pb, 1, 1); /* vop coded */ if ( s->pict_type == P_TYPE || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE)) { - put_bits(&s->pb, 1, s->no_rounding); /* rounding type */ - } - put_bits(&s->pb, 3, 0); /* intra dc VLC threshold */ + put_bits(&s->pb, 1, s->no_rounding); /* rounding type */ + } + put_bits(&s->pb, 3, 0); /* intra dc VLC threshold */ if(!s->progressive_sequence){ put_bits(&s->pb, 1, s->current_picture_ptr->top_field_first); put_bits(&s->pb, 1, s->alternate_scan); @@ -2453,9 +2453,9 @@ put_bits(&s->pb, 5, s->qscale); if (s->pict_type != I_TYPE) - put_bits(&s->pb, 3, s->f_code); /* fcode_for */ + put_bits(&s->pb, 3, s->f_code); /* fcode_for */ if (s->pict_type == B_TYPE) - put_bits(&s->pb, 3, s->b_code); /* fcode_back */ + put_bits(&s->pb, 3, s->b_code); /* fcode_back */ // printf("****frame %d\n", picture_number); } @@ -2492,9 +2492,9 @@ /* find prediction */ if (n < 4) { - scale = s->y_dc_scale; + scale = s->y_dc_scale; } else { - scale = s->c_dc_scale; + scale = s->c_dc_scale; } if(IS_3IV1) scale= 8; @@ -2520,10 +2520,10 @@ } if (abs(a - b) < abs(b - c)) { - pred = c; + pred = c; *dir_ptr = 1; /* top */ } else { - pred = a; + pred = a; *dir_ptr = 0; /* left */ } /* we assume pred is positive */ @@ -2629,11 +2629,11 @@ // if(level<-255 || level>255) printf("dc overflow\n"); level+=256; if (n < 4) { - /* luminance */ - put_bits(s, uni_DCtab_lum_len[level], uni_DCtab_lum_bits[level]); + /* luminance */ + put_bits(s, uni_DCtab_lum_len[level], uni_DCtab_lum_bits[level]); } else { - /* chrominance */ - put_bits(s, uni_DCtab_chrom_len[level], uni_DCtab_chrom_bits[level]); + /* chrominance */ + put_bits(s, uni_DCtab_chrom_len[level], uni_DCtab_chrom_bits[level]); } #else int size, v; @@ -2641,25 +2641,25 @@ size = 0; v = abs(level); while (v) { - v >>= 1; - size++; + v >>= 1; + size++; } if (n < 4) { - /* luminance */ - put_bits(&s->pb, DCtab_lum[size][1], DCtab_lum[size][0]); + /* luminance */ + put_bits(&s->pb, DCtab_lum[size][1], DCtab_lum[size][0]); } else { - /* chrominance */ - put_bits(&s->pb, DCtab_chrom[size][1], DCtab_chrom[size][0]); + /* chrominance */ + put_bits(&s->pb, DCtab_chrom[size][1], DCtab_chrom[size][0]); } /* encode remaining bits */ if (size > 0) { - if (level < 0) - level = (-level) ^ ((1 << size) - 1); - put_bits(&s->pb, size, level); - if (size > 8) - put_bits(&s->pb, 1, 1); + if (level < 0) + level = (-level) ^ ((1 << size) - 1); + put_bits(&s->pb, size, level); + if (size > 8) + put_bits(&s->pb, 1, 1); } #endif } @@ -2689,16 +2689,16 @@ const int last_index = s->block_last_index[n]; if (s->mb_intra) { //Note gcc (3.2.1 at least) will optimize this away - /* mpeg4 based DC predictor */ - mpeg4_encode_dc(dc_pb, intra_dc, n); + /* mpeg4 based DC predictor */ + mpeg4_encode_dc(dc_pb, intra_dc, n); if(last_index<1) return; - i = 1; + i = 1; rl = &rl_intra; bits_tab= uni_mpeg4_intra_rl_bits; len_tab = uni_mpeg4_intra_rl_len; } else { if(last_index<0) return; - i = 0; + i = 0; rl = &rl_inter; bits_tab= uni_mpeg4_inter_rl_bits; len_tab = uni_mpeg4_inter_rl_len; @@ -2708,9 +2708,9 @@ last_non_zero = i - 1; #if 1 for (; i < last_index; i++) { - int level = block[ scan_table[i] ]; - if (level) { - int run = i - last_non_zero - 1; + int level = block[ scan_table[i] ]; + if (level) { + int run = i - last_non_zero - 1; level+=64; if((level&(~127)) == 0){ const int index= UNI_MPEG4_ENC_INDEX(0, run, level); @@ -2718,11 +2718,11 @@ }else{ //ESC3 put_bits(ac_pb, 7+2+1+6+1+12+1, (3<<23)+(3<<21)+(0<<20)+(run<<14)+(1<<13)+(((level-64)&0xfff)<<1)+1); } - last_non_zero = i; - } + last_non_zero = i; + } } /*if(i<=last_index)*/{ - int level = block[ scan_table[i] ]; + int level = block[ scan_table[i] ]; int run = i - last_non_zero - 1; level+=64; if((level&(~127)) == 0){ @@ -2734,17 +2734,17 @@ } #else for (; i <= last_index; i++) { - const int slevel = block[ scan_table[i] ]; - if (slevel) { + const int slevel = block[ scan_table[i] ]; + if (slevel) { int level; - int run = i - last_non_zero - 1; - last = (i == last_index); - sign = 0; - level = slevel; - if (level < 0) { - sign = 1; - level = -level; - } + int run = i - last_non_zero - 1; + last = (i == last_index); + sign = 0; + level = slevel; + if (level < 0) { + sign = 1; + level = -level; + } code = get_rl_index(rl, last, run, level); put_bits(ac_pb, rl->table_vlc[code][1], rl->table_vlc[code][0]); if (code == rl->n) { @@ -2786,8 +2786,8 @@ } else { put_bits(ac_pb, 1, sign); } - last_non_zero = i; - } + last_non_zero = i; + } } #endif } @@ -2802,15 +2802,15 @@ int len=0; if (s->mb_intra) { //Note gcc (3.2.1 at least) will optimize this away - /* mpeg4 based DC predictor */ - len += mpeg4_get_dc_length(intra_dc, n); + /* mpeg4 based DC predictor */ + len += mpeg4_get_dc_length(intra_dc, n); if(last_index<1) return len; - i = 1; + i = 1; rl = &rl_intra; len_tab = uni_mpeg4_intra_rl_len; } else { if(last_index<0) return 0; - i = 0; + i = 0; rl = &rl_inter; len_tab = uni_mpeg4_inter_rl_len; } @@ -2818,9 +2818,9 @@ /* AC coefs */ last_non_zero = i - 1; for (; i < last_index; i++) { - int level = block[ scan_table[i] ]; - if (level) { - int run = i - last_non_zero - 1; + int level = block[ scan_table[i] ]; + if (level) { + int run = i - last_non_zero - 1; level+=64; if((level&(~127)) == 0){ const int index= UNI_MPEG4_ENC_INDEX(0, run, level); @@ -2828,11 +2828,11 @@ }else{ //ESC3 len += 7+2+1+6+1+12+1; } - last_non_zero = i; - } + last_non_zero = i; + } } /*if(i<=last_index)*/{ - int level = block[ scan_table[i] ]; + int level = block[ scan_table[i] ]; int run = i - last_non_zero - 1; level+=64; if((level&(~127)) == 0){ @@ -3251,7 +3251,7 @@ //FIXME reduced res stuff here if (s->pict_type != I_TYPE) { - int f_code = get_bits(&s->gb, 3); /* fcode_for */ + int f_code = get_bits(&s->gb, 3); /* fcode_for */ if(f_code==0){ av_log(s->avctx, AV_LOG_ERROR, "Error, video packet header damaged (f_code=0)\n"); } @@ -4741,7 +4741,7 @@ if(intra) { if(s->qscale < s->intra_dc_threshold){ - /* DC coef */ + /* DC coef */ if(s->partitioned_frame){ level = s->dc_val[0][ s->block_index[n] ]; if(n<4) level= FASTDIV((level + (s->y_dc_scale>>1)), s->y_dc_scale); @@ -4898,7 +4898,7 @@ } } #endif - if (level>0) level= level * qmul + qadd; + if (level>0) level= level * qmul + qadd; else level= level * qmul - qadd; if((unsigned)(level + 2048) > 4095){ @@ -5014,18 +5014,18 @@ } if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "Bad H263 id\n"); - return -1; /* h263 id */ - } - skip_bits1(&s->gb); /* split screen off */ - skip_bits1(&s->gb); /* camera off */ - skip_bits1(&s->gb); /* freeze picture release off */ + return -1; /* h263 id */ + } + skip_bits1(&s->gb); /* split screen off */ + skip_bits1(&s->gb); /* camera off */ + skip_bits1(&s->gb); /* freeze picture release off */ format = get_bits(&s->gb, 3); /* 0 forbidden 1 sub-QCIF 10 QCIF - 7 extended PTYPE (PLUSPTYPE) + 7 extended PTYPE (PLUSPTYPE) */ if (format != 7 && format != 6) { @@ -5042,17 +5042,17 @@ if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "H263 SAC not supported\n"); - return -1; /* SAC: off */ + return -1; /* SAC: off */ } s->obmc= get_bits1(&s->gb); /* Advanced prediction mode */ s->unrestricted_mv = s->h263_long_vectors || s->obmc; if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "H263 PB frame not supported\n"); - return -1; /* not PB frame */ + return -1; /* not PB frame */ } s->chroma_qscale= s->qscale = get_bits(&s->gb, 5); - skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */ + skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */ s->width = width; s->height = height; @@ -5511,17 +5511,17 @@ } s->low_delay= get_bits1(gb); if(get_bits1(gb)){ /* vbv parameters */ - get_bits(gb, 15); /* first_half_bitrate */ - skip_bits1(gb); /* marker */ - get_bits(gb, 15); /* latter_half_bitrate */ - skip_bits1(gb); /* marker */ - get_bits(gb, 15); /* first_half_vbv_buffer_size */ - skip_bits1(gb); /* marker */ - get_bits(gb, 3); /* latter_half_vbv_buffer_size */ - get_bits(gb, 11); /* first_half_vbv_occupancy */ - skip_bits1(gb); /* marker */ - get_bits(gb, 15); /* latter_half_vbv_occupancy */ - skip_bits1(gb); /* marker */ + get_bits(gb, 15); /* first_half_bitrate */ + skip_bits1(gb); /* marker */ + get_bits(gb, 15); /* latter_half_bitrate */ + skip_bits1(gb); /* marker */ + get_bits(gb, 15); /* first_half_vbv_buffer_size */ + skip_bits1(gb); /* marker */ + get_bits(gb, 3); /* latter_half_vbv_buffer_size */ + get_bits(gb, 11); /* first_half_vbv_occupancy */ + skip_bits1(gb); /* marker */ + get_bits(gb, 15); /* latter_half_vbv_occupancy */ + skip_bits1(gb); /* marker */ } }else{ // set low delay flag only once the smartest? low delay detection won't be overriden @@ -5628,7 +5628,7 @@ /* load custom intra matrix */ if(get_bits1(gb)){ int last=0; - for(i=0; i<64; i++){ + for(i=0; i<64; i++){ int j; v= get_bits(gb, 8); if(v==0) break; @@ -5641,7 +5641,7 @@ /* replicate last value */ for(; i<64; i++){ - int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ]; + int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ]; s->intra_matrix[j]= last; s->chroma_intra_matrix[j]= last; } @@ -5650,7 +5650,7 @@ /* load custom non intra matrix */ if(get_bits1(gb)){ int last=0; - for(i=0; i<64; i++){ + for(i=0; i<64; i++){ int j; v= get_bits(gb, 8); if(v==0) break; @@ -5663,7 +5663,7 @@ /* replicate last value */ for(; i<64; i++){ - int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ]; + int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ]; s->inter_matrix[j]= last; s->chroma_inter_matrix[j]= last; } @@ -5794,7 +5794,7 @@ static int decode_vop_header(MpegEncContext *s, GetBitContext *gb){ int time_incr, time_increment; - s->pict_type = get_bits(gb, 2) + I_TYPE; /* pict type: I = 0 , P = 1 */ + s->pict_type = get_bits(gb, 2) + I_TYPE; /* pict type: I = 0 , P = 1 */ if(s->pict_type==B_TYPE && s->low_delay && s->vol_control_parameters==0 && !(s->flags & CODEC_FLAG_LOW_DELAY)){ av_log(s->avctx, AV_LOG_ERROR, "low_delay flag incorrectly, clearing it\n"); s->low_delay=0; @@ -5877,9 +5877,9 @@ if (s->shape != BIN_ONLY_SHAPE && ( s->pict_type == P_TYPE || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE))) { /* rounding type for motion estimation */ - s->no_rounding = get_bits1(gb); + s->no_rounding = get_bits1(gb); } else { - s->no_rounding = 0; + s->no_rounding = 0; } //FIXME reduced res stuff @@ -5938,7 +5938,7 @@ } if (s->pict_type != I_TYPE) { - s->f_code = get_bits(gb, 3); /* fcode_for */ + s->f_code = get_bits(gb, 3); /* fcode_for */ if(s->f_code==0){ av_log(s->avctx, AV_LOG_ERROR, "Error, header damaged or not MPEG4 header (f_code=0)\n"); return -1; // makes no sense to continue, as the MV decoding will break very quickly @@ -6094,15 +6094,15 @@ if (get_bits1(&s->gb) != 1) { av_log(s->avctx, AV_LOG_ERROR, "Bad marker\n"); - return -1; /* marker */ + return -1; /* marker */ } if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "Bad H263 id\n"); - return -1; /* h263 id */ - } - skip_bits1(&s->gb); /* split screen off */ - skip_bits1(&s->gb); /* camera off */ - skip_bits1(&s->gb); /* freeze picture release off */ + return -1; /* h263 id */ + } + skip_bits1(&s->gb); /* split screen off */ + skip_bits1(&s->gb); /* camera off */ + skip_bits1(&s->gb); /* freeze picture release off */ format = get_bits(&s->gb, 3); if (format != 7) { @@ -6118,23 +6118,23 @@ if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "SAC not supported\n"); - return -1; /* SAC: off */ + return -1; /* SAC: off */ } if (get_bits1(&s->gb) != 0) { s->obmc= 1; av_log(s->avctx, AV_LOG_ERROR, "Advanced Prediction Mode not supported\n"); -// return -1; /* advanced prediction mode: off */ +// return -1; /* advanced prediction mode: off */ } if (get_bits1(&s->gb) != 0) { av_log(s->avctx, AV_LOG_ERROR, "PB frame mode no supported\n"); - return -1; /* PB frame mode */ + return -1; /* PB frame mode */ } /* skip unknown header garbage */ skip_bits(&s->gb, 41); s->chroma_qscale= s->qscale = get_bits(&s->gb, 5); - skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */ + skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */ /* PEI */ while (get_bits1(&s->gb) != 0) { @@ -6208,7 +6208,7 @@ if (s->dropable) s->pict_type = P_TYPE; - skip_bits1(&s->gb); /* deblocking flag */ + skip_bits1(&s->gb); /* deblocking flag */ s->chroma_qscale= s->qscale = get_bits(&s->gb, 5); s->h263_plus = 0;
--- a/h264.c Wed Dec 21 17:50:40 2005 +0000 +++ b/h264.c Thu Dec 22 01:10:11 2005 +0000 @@ -147,15 +147,15 @@ MpegEncContext s; int nal_ref_idc; int nal_unit_type; -#define NAL_SLICE 1 -#define NAL_DPA 2 -#define NAL_DPB 3 -#define NAL_DPC 4 -#define NAL_IDR_SLICE 5 -#define NAL_SEI 6 -#define NAL_SPS 7 -#define NAL_PPS 8 -#define NAL_AUD 9 +#define NAL_SLICE 1 +#define NAL_DPA 2 +#define NAL_DPB 3 +#define NAL_DPC 4 +#define NAL_IDR_SLICE 5 +#define NAL_SEI 6 +#define NAL_SPS 7 +#define NAL_PPS 8 +#define NAL_AUD 9 #define NAL_END_SEQUENCE 10 #define NAL_END_STREAM 11 #define NAL_FILLER_DATA 12 @@ -1461,7 +1461,7 @@ int i, si, di; uint8_t *dst; -// src[0]&0x80; //forbidden bit +// src[0]&0x80; //forbidden bit h->nal_ref_idc= src[0]>>5; h->nal_unit_type= src[0]&0x1F; @@ -7545,8 +7545,8 @@ case NAL_SPS_EXT: case NAL_AUXILIARY_SLICE: break; - default: - av_log(avctx, AV_LOG_ERROR, "Unknown NAL code: %d\n", h->nal_unit_type); + default: + av_log(avctx, AV_LOG_ERROR, "Unknown NAL code: %d\n", h->nal_unit_type); } }
--- a/i386/cputest.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/cputest.c Thu Dec 22 01:10:11 2005 +0000 @@ -15,7 +15,7 @@ /* ebx saving is necessary for PIC. gcc seems unable to see it alone */ #define cpuid(index,eax,ebx,ecx,edx)\ __asm __volatile\ - ("mov %%"REG_b", %%"REG_S"\n\t"\ + ("mov %%"REG_b", %%"REG_S"\n\t"\ "cpuid\n\t"\ "xchg %%"REG_b", %%"REG_S\ : "=a" (eax), "=S" (ebx),\ @@ -89,8 +89,8 @@ edx == 0x48727561 && ecx == 0x736c7561) { /* "CentaurHauls" */ /* VIA C3 */ - if(ext_caps & (1<<24)) - rval |= MM_MMXEXT; + if(ext_caps & (1<<24)) + rval |= MM_MMXEXT; } else if (ebx == 0x69727943 && edx == 0x736e4978 && ecx == 0x64616574) {
--- a/i386/dsputil_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/dsputil_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -89,56 +89,56 @@ // first argument is unmodifed and second is trashed // regfe is supposed to contain 0xfefefefefefefefe #define PAVGB_MMX_NO_RND(rega, regb, regr, regfe) \ - "movq " #rega ", " #regr " \n\t"\ - "pand " #regb ", " #regr " \n\t"\ - "pxor " #rega ", " #regb " \n\t"\ - "pand " #regfe "," #regb " \n\t"\ - "psrlq $1, " #regb " \n\t"\ - "paddb " #regb ", " #regr " \n\t" + "movq " #rega ", " #regr " \n\t"\ + "pand " #regb ", " #regr " \n\t"\ + "pxor " #rega ", " #regb " \n\t"\ + "pand " #regfe "," #regb " \n\t"\ + "psrlq $1, " #regb " \n\t"\ + "paddb " #regb ", " #regr " \n\t" #define PAVGB_MMX(rega, regb, regr, regfe) \ - "movq " #rega ", " #regr " \n\t"\ - "por " #regb ", " #regr " \n\t"\ - "pxor " #rega ", " #regb " \n\t"\ - "pand " #regfe "," #regb " \n\t"\ - "psrlq $1, " #regb " \n\t"\ - "psubb " #regb ", " #regr " \n\t" + "movq " #rega ", " #regr " \n\t"\ + "por " #regb ", " #regr " \n\t"\ + "pxor " #rega ", " #regb " \n\t"\ + "pand " #regfe "," #regb " \n\t"\ + "psrlq $1, " #regb " \n\t"\ + "psubb " #regb ", " #regr " \n\t" // mm6 is supposed to contain 0xfefefefefefefefe #define PAVGBP_MMX_NO_RND(rega, regb, regr, regc, regd, regp) \ - "movq " #rega ", " #regr " \n\t"\ - "movq " #regc ", " #regp " \n\t"\ - "pand " #regb ", " #regr " \n\t"\ - "pand " #regd ", " #regp " \n\t"\ - "pxor " #rega ", " #regb " \n\t"\ - "pxor " #regc ", " #regd " \n\t"\ - "pand %%mm6, " #regb " \n\t"\ - "pand %%mm6, " #regd " \n\t"\ - "psrlq $1, " #regb " \n\t"\ - "psrlq $1, " #regd " \n\t"\ - "paddb " #regb ", " #regr " \n\t"\ - "paddb " #regd ", " #regp " \n\t" + "movq " #rega ", " #regr " \n\t"\ + "movq " #regc ", " #regp " \n\t"\ + "pand " #regb ", " #regr " \n\t"\ + "pand " #regd ", " #regp " \n\t"\ + "pxor " #rega ", " #regb " \n\t"\ + "pxor " #regc ", " #regd " \n\t"\ + "pand %%mm6, " #regb " \n\t"\ + "pand %%mm6, " #regd " \n\t"\ + "psrlq $1, " #regb " \n\t"\ + "psrlq $1, " #regd " \n\t"\ + "paddb " #regb ", " #regr " \n\t"\ + "paddb " #regd ", " #regp " \n\t" #define PAVGBP_MMX(rega, regb, regr, regc, regd, regp) \ - "movq " #rega ", " #regr " \n\t"\ - "movq " #regc ", " #regp " \n\t"\ - "por " #regb ", " #regr " \n\t"\ - "por " #regd ", " #regp " \n\t"\ - "pxor " #rega ", " #regb " \n\t"\ - "pxor " #regc ", " #regd " \n\t"\ - "pand %%mm6, " #regb " \n\t"\ - "pand %%mm6, " #regd " \n\t"\ - "psrlq $1, " #regd " \n\t"\ - "psrlq $1, " #regb " \n\t"\ - "psubb " #regb ", " #regr " \n\t"\ - "psubb " #regd ", " #regp " \n\t" + "movq " #rega ", " #regr " \n\t"\ + "movq " #regc ", " #regp " \n\t"\ + "por " #regb ", " #regr " \n\t"\ + "por " #regd ", " #regp " \n\t"\ + "pxor " #rega ", " #regb " \n\t"\ + "pxor " #regc ", " #regd " \n\t"\ + "pand %%mm6, " #regb " \n\t"\ + "pand %%mm6, " #regd " \n\t"\ + "psrlq $1, " #regd " \n\t"\ + "psrlq $1, " #regb " \n\t"\ + "psubb " #regb ", " #regr " \n\t"\ + "psubb " #regd ", " #regp " \n\t" /***********************************/ /* MMX no rounding */ #define DEF(x, y) x ## _no_rnd_ ## y ##_mmx #define SET_RND MOVQ_WONE -#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX_NO_RND(a, b, c, d, e, f) -#define PAVGB(a, b, c, e) PAVGB_MMX_NO_RND(a, b, c, e) +#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX_NO_RND(a, b, c, d, e, f) +#define PAVGB(a, b, c, e) PAVGB_MMX_NO_RND(a, b, c, e) #include "dsputil_mmx_rnd.h" @@ -151,8 +151,8 @@ #define DEF(x, y) x ## _ ## y ##_mmx #define SET_RND MOVQ_WTWO -#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX(a, b, c, d, e, f) -#define PAVGB(a, b, c, e) PAVGB_MMX(a, b, c, e) +#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX(a, b, c, d, e, f) +#define PAVGB(a, b, c, e) PAVGB_MMX(a, b, c, e) #include "dsputil_mmx_rnd.h" @@ -193,25 +193,25 @@ static void get_pixels_mmx(DCTELEM *block, const uint8_t *pixels, int line_size) { asm volatile( - "mov $-128, %%"REG_a" \n\t" - "pxor %%mm7, %%mm7 \n\t" - ".balign 16 \n\t" - "1: \n\t" - "movq (%0), %%mm0 \n\t" - "movq (%0, %2), %%mm2 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "movq %%mm0, (%1, %%"REG_a")\n\t" - "movq %%mm1, 8(%1, %%"REG_a")\n\t" - "movq %%mm2, 16(%1, %%"REG_a")\n\t" - "movq %%mm3, 24(%1, %%"REG_a")\n\t" - "add %3, %0 \n\t" - "add $32, %%"REG_a" \n\t" - "js 1b \n\t" + "mov $-128, %%"REG_a" \n\t" + "pxor %%mm7, %%mm7 \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0), %%mm0 \n\t" + "movq (%0, %2), %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "movq %%mm0, (%1, %%"REG_a") \n\t" + "movq %%mm1, 8(%1, %%"REG_a") \n\t" + "movq %%mm2, 16(%1, %%"REG_a") \n\t" + "movq %%mm3, 24(%1, %%"REG_a") \n\t" + "add %3, %0 \n\t" + "add $32, %%"REG_a" \n\t" + "js 1b \n\t" : "+r" (pixels) : "r" (block+64), "r" ((long)line_size), "r" ((long)line_size*2) : "%"REG_a @@ -221,26 +221,26 @@ static inline void diff_pixels_mmx(DCTELEM *block, const uint8_t *s1, const uint8_t *s2, int stride) { asm volatile( - "pxor %%mm7, %%mm7 \n\t" - "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" - "1: \n\t" - "movq (%0), %%mm0 \n\t" - "movq (%1), %%mm2 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "psubw %%mm2, %%mm0 \n\t" - "psubw %%mm3, %%mm1 \n\t" - "movq %%mm0, (%2, %%"REG_a")\n\t" - "movq %%mm1, 8(%2, %%"REG_a")\n\t" - "add %3, %0 \n\t" - "add %3, %1 \n\t" - "add $16, %%"REG_a" \n\t" - "jnz 1b \n\t" + "pxor %%mm7, %%mm7 \n\t" + "mov $-128, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0), %%mm0 \n\t" + "movq (%1), %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "movq %%mm0, (%2, %%"REG_a") \n\t" + "movq %%mm1, 8(%2, %%"REG_a") \n\t" + "add %3, %0 \n\t" + "add %3, %1 \n\t" + "add $16, %%"REG_a" \n\t" + "jnz 1b \n\t" : "+r" (s1), "+r" (s2) : "r" (block+64), "r" ((long)stride) : "%"REG_a @@ -257,25 +257,25 @@ p = block; pix = pixels; /* unrolled loop */ - __asm __volatile( - "movq %3, %%mm0\n\t" - "movq 8%3, %%mm1\n\t" - "movq 16%3, %%mm2\n\t" - "movq 24%3, %%mm3\n\t" - "movq 32%3, %%mm4\n\t" - "movq 40%3, %%mm5\n\t" - "movq 48%3, %%mm6\n\t" - "movq 56%3, %%mm7\n\t" - "packuswb %%mm1, %%mm0\n\t" - "packuswb %%mm3, %%mm2\n\t" - "packuswb %%mm5, %%mm4\n\t" - "packuswb %%mm7, %%mm6\n\t" - "movq %%mm0, (%0)\n\t" - "movq %%mm2, (%0, %1)\n\t" - "movq %%mm4, (%0, %1, 2)\n\t" - "movq %%mm6, (%0, %2)\n\t" - ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "m"(*p) - :"memory"); + __asm __volatile( + "movq %3, %%mm0 \n\t" + "movq 8%3, %%mm1 \n\t" + "movq 16%3, %%mm2 \n\t" + "movq 24%3, %%mm3 \n\t" + "movq 32%3, %%mm4 \n\t" + "movq 40%3, %%mm5 \n\t" + "movq 48%3, %%mm6 \n\t" + "movq 56%3, %%mm7 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "packuswb %%mm3, %%mm2 \n\t" + "packuswb %%mm5, %%mm4 \n\t" + "packuswb %%mm7, %%mm6 \n\t" + "movq %%mm0, (%0) \n\t" + "movq %%mm2, (%0, %1) \n\t" + "movq %%mm4, (%0, %1, 2) \n\t" + "movq %%mm6, (%0, %2) \n\t" + ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "m"(*p) + :"memory"); pix += line_size*4; p += 32; @@ -283,24 +283,24 @@ // compiler would generate some very strange code // thus using "r" __asm __volatile( - "movq (%3), %%mm0\n\t" - "movq 8(%3), %%mm1\n\t" - "movq 16(%3), %%mm2\n\t" - "movq 24(%3), %%mm3\n\t" - "movq 32(%3), %%mm4\n\t" - "movq 40(%3), %%mm5\n\t" - "movq 48(%3), %%mm6\n\t" - "movq 56(%3), %%mm7\n\t" - "packuswb %%mm1, %%mm0\n\t" - "packuswb %%mm3, %%mm2\n\t" - "packuswb %%mm5, %%mm4\n\t" - "packuswb %%mm7, %%mm6\n\t" - "movq %%mm0, (%0)\n\t" - "movq %%mm2, (%0, %1)\n\t" - "movq %%mm4, (%0, %1, 2)\n\t" - "movq %%mm6, (%0, %2)\n\t" - ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "r"(p) - :"memory"); + "movq (%3), %%mm0 \n\t" + "movq 8(%3), %%mm1 \n\t" + "movq 16(%3), %%mm2 \n\t" + "movq 24(%3), %%mm3 \n\t" + "movq 32(%3), %%mm4 \n\t" + "movq 40(%3), %%mm5 \n\t" + "movq 48(%3), %%mm6 \n\t" + "movq 56(%3), %%mm7 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "packuswb %%mm3, %%mm2 \n\t" + "packuswb %%mm5, %%mm4 \n\t" + "packuswb %%mm7, %%mm6 \n\t" + "movq %%mm0, (%0) \n\t" + "movq %%mm2, (%0, %1) \n\t" + "movq %%mm4, (%0, %1, 2) \n\t" + "movq %%mm6, (%0, %2) \n\t" + ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "r"(p) + :"memory"); } static const unsigned char __align8 vector128[8] = @@ -333,30 +333,30 @@ MOVQ_ZERO(mm7); i = 4; do { - __asm __volatile( - "movq (%2), %%mm0\n\t" - "movq 8(%2), %%mm1\n\t" - "movq 16(%2), %%mm2\n\t" - "movq 24(%2), %%mm3\n\t" - "movq %0, %%mm4\n\t" - "movq %1, %%mm6\n\t" - "movq %%mm4, %%mm5\n\t" - "punpcklbw %%mm7, %%mm4\n\t" - "punpckhbw %%mm7, %%mm5\n\t" - "paddsw %%mm4, %%mm0\n\t" - "paddsw %%mm5, %%mm1\n\t" - "movq %%mm6, %%mm5\n\t" - "punpcklbw %%mm7, %%mm6\n\t" - "punpckhbw %%mm7, %%mm5\n\t" - "paddsw %%mm6, %%mm2\n\t" - "paddsw %%mm5, %%mm3\n\t" - "packuswb %%mm1, %%mm0\n\t" - "packuswb %%mm3, %%mm2\n\t" - "movq %%mm0, %0\n\t" - "movq %%mm2, %1\n\t" - :"+m"(*pix), "+m"(*(pix+line_size)) - :"r"(p) - :"memory"); + __asm __volatile( + "movq (%2), %%mm0 \n\t" + "movq 8(%2), %%mm1 \n\t" + "movq 16(%2), %%mm2 \n\t" + "movq 24(%2), %%mm3 \n\t" + "movq %0, %%mm4 \n\t" + "movq %1, %%mm6 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddsw %%mm4, %%mm0 \n\t" + "paddsw %%mm5, %%mm1 \n\t" + "movq %%mm6, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm6 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddsw %%mm6, %%mm2 \n\t" + "paddsw %%mm5, %%mm3 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "packuswb %%mm3, %%mm2 \n\t" + "movq %%mm0, %0 \n\t" + "movq %%mm2, %1 \n\t" + :"+m"(*pix), "+m"(*(pix+line_size)) + :"r"(p) + :"memory"); pix += line_size*2; p += 16; } while (--i); @@ -365,101 +365,101 @@ static void put_pixels4_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movd (%1), %%mm0 \n\t" - "movd (%1, %3), %%mm1 \n\t" - "movd %%mm0, (%2) \n\t" - "movd %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movd (%1), %%mm0 \n\t" - "movd (%1, %3), %%mm1 \n\t" - "movd %%mm0, (%2) \n\t" - "movd %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - : "+g"(h), "+r" (pixels), "+r" (block) - : "r"((long)line_size) - : "%"REG_a, "memory" - ); + "lea (%3, %3), %%"REG_a" \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movd (%1), %%mm0 \n\t" + "movd (%1, %3), %%mm1 \n\t" + "movd %%mm0, (%2) \n\t" + "movd %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movd (%1), %%mm0 \n\t" + "movd (%1, %3), %%mm1 \n\t" + "movd %%mm0, (%2) \n\t" + "movd %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + : "+g"(h), "+r" (pixels), "+r" (block) + : "r"((long)line_size) + : "%"REG_a, "memory" + ); } static void put_pixels8_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - : "+g"(h), "+r" (pixels), "+r" (block) - : "r"((long)line_size) - : "%"REG_a, "memory" - ); + "lea (%3, %3), %%"REG_a" \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + : "+g"(h), "+r" (pixels), "+r" (block) + : "r"((long)line_size) + : "%"REG_a, "memory" + ); } static void put_pixels16_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm4 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq 8(%1, %3), %%mm5 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm4, 8(%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "movq %%mm5, 8(%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm4 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq 8(%1, %3), %%mm5 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm4, 8(%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "movq %%mm5, 8(%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - : "+g"(h), "+r" (pixels), "+r" (block) - : "r"((long)line_size) - : "%"REG_a, "memory" - ); + "lea (%3, %3), %%"REG_a" \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm4 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq 8(%1, %3), %%mm5 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm4, 8(%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "movq %%mm5, 8(%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm4 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq 8(%1, %3), %%mm5 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm4, 8(%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "movq %%mm5, 8(%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + : "+g"(h), "+r" (pixels), "+r" (block) + : "r"((long)line_size) + : "%"REG_a, "memory" + ); } static void clear_blocks_mmx(DCTELEM *blocks) { __asm __volatile( - "pxor %%mm7, %%mm7 \n\t" - "mov $-128*6, %%"REG_a" \n\t" - "1: \n\t" - "movq %%mm7, (%0, %%"REG_a") \n\t" - "movq %%mm7, 8(%0, %%"REG_a") \n\t" - "movq %%mm7, 16(%0, %%"REG_a") \n\t" - "movq %%mm7, 24(%0, %%"REG_a") \n\t" - "add $32, %%"REG_a" \n\t" - " js 1b \n\t" + "pxor %%mm7, %%mm7 \n\t" + "mov $-128*6, %%"REG_a" \n\t" + "1: \n\t" + "movq %%mm7, (%0, %%"REG_a") \n\t" + "movq %%mm7, 8(%0, %%"REG_a") \n\t" + "movq %%mm7, 16(%0, %%"REG_a") \n\t" + "movq %%mm7, 24(%0, %%"REG_a") \n\t" + "add $32, %%"REG_a" \n\t" + " js 1b \n\t" : : "r" (((uint8_t *)blocks)+128*6) : "%"REG_a ); @@ -472,31 +472,31 @@ long index= -line_size*h; __asm __volatile( - "pxor %%mm7, %%mm7 \n\t" - "pxor %%mm6, %%mm6 \n\t" - "1: \n\t" - "movq (%2, %1), %%mm0 \n\t" - "movq (%2, %1), %%mm1 \n\t" - "movq 8(%2, %1), %%mm2 \n\t" - "movq 8(%2, %1), %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "paddw %%mm0, %%mm1 \n\t" - "paddw %%mm2, %%mm3 \n\t" - "paddw %%mm1, %%mm3 \n\t" - "paddw %%mm3, %%mm6 \n\t" - "add %3, %1 \n\t" - " js 1b \n\t" - "movq %%mm6, %%mm5 \n\t" - "psrlq $32, %%mm6 \n\t" - "paddw %%mm5, %%mm6 \n\t" - "movq %%mm6, %%mm5 \n\t" - "psrlq $16, %%mm6 \n\t" - "paddw %%mm5, %%mm6 \n\t" - "movd %%mm6, %0 \n\t" - "andl $0xFFFF, %0 \n\t" + "pxor %%mm7, %%mm7 \n\t" + "pxor %%mm6, %%mm6 \n\t" + "1: \n\t" + "movq (%2, %1), %%mm0 \n\t" + "movq (%2, %1), %%mm1 \n\t" + "movq 8(%2, %1), %%mm2 \n\t" + "movq 8(%2, %1), %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "paddw %%mm0, %%mm1 \n\t" + "paddw %%mm2, %%mm3 \n\t" + "paddw %%mm1, %%mm3 \n\t" + "paddw %%mm3, %%mm6 \n\t" + "add %3, %1 \n\t" + " js 1b \n\t" + "movq %%mm6, %%mm5 \n\t" + "psrlq $32, %%mm6 \n\t" + "paddw %%mm5, %%mm6 \n\t" + "movq %%mm6, %%mm5 \n\t" + "psrlq $16, %%mm6 \n\t" + "paddw %%mm5, %%mm6 \n\t" + "movd %%mm6, %0 \n\t" + "andl $0xFFFF, %0 \n\t" : "=&r" (sum), "+r" (index) : "r" (pix - index), "r" ((long)line_size) ); @@ -508,18 +508,18 @@ static void add_bytes_mmx(uint8_t *dst, uint8_t *src, int w){ long i=0; asm volatile( - "1: \n\t" - "movq (%1, %0), %%mm0 \n\t" - "movq (%2, %0), %%mm1 \n\t" - "paddb %%mm0, %%mm1 \n\t" - "movq %%mm1, (%2, %0) \n\t" - "movq 8(%1, %0), %%mm0 \n\t" - "movq 8(%2, %0), %%mm1 \n\t" - "paddb %%mm0, %%mm1 \n\t" - "movq %%mm1, 8(%2, %0) \n\t" - "add $16, %0 \n\t" - "cmp %3, %0 \n\t" - " jb 1b \n\t" + "1: \n\t" + "movq (%1, %0), %%mm0 \n\t" + "movq (%2, %0), %%mm1 \n\t" + "paddb %%mm0, %%mm1 \n\t" + "movq %%mm1, (%2, %0) \n\t" + "movq 8(%1, %0), %%mm0 \n\t" + "movq 8(%2, %0), %%mm1 \n\t" + "paddb %%mm0, %%mm1 \n\t" + "movq %%mm1, 8(%2, %0) \n\t" + "add $16, %0 \n\t" + "cmp %3, %0 \n\t" + " jb 1b \n\t" : "+r" (i) : "r"(src), "r"(dst), "r"((long)w-15) ); @@ -528,75 +528,75 @@ } #define H263_LOOP_FILTER \ - "pxor %%mm7, %%mm7 \n\t"\ - "movq %0, %%mm0 \n\t"\ - "movq %0, %%mm1 \n\t"\ - "movq %3, %%mm2 \n\t"\ - "movq %3, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "psubw %%mm2, %%mm0 \n\t"\ - "psubw %%mm3, %%mm1 \n\t"\ - "movq %1, %%mm2 \n\t"\ - "movq %1, %%mm3 \n\t"\ - "movq %2, %%mm4 \n\t"\ - "movq %2, %%mm5 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ - "punpckhbw %%mm7, %%mm5 \n\t"\ - "psubw %%mm2, %%mm4 \n\t"\ - "psubw %%mm3, %%mm5 \n\t"\ - "psllw $2, %%mm4 \n\t"\ - "psllw $2, %%mm5 \n\t"\ - "paddw %%mm0, %%mm4 \n\t"\ - "paddw %%mm1, %%mm5 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "pcmpgtw %%mm4, %%mm6 \n\t"\ - "pcmpgtw %%mm5, %%mm7 \n\t"\ - "pxor %%mm6, %%mm4 \n\t"\ - "pxor %%mm7, %%mm5 \n\t"\ - "psubw %%mm6, %%mm4 \n\t"\ - "psubw %%mm7, %%mm5 \n\t"\ - "psrlw $3, %%mm4 \n\t"\ - "psrlw $3, %%mm5 \n\t"\ - "packuswb %%mm5, %%mm4 \n\t"\ - "packsswb %%mm7, %%mm6 \n\t"\ - "pxor %%mm7, %%mm7 \n\t"\ - "movd %4, %%mm2 \n\t"\ - "punpcklbw %%mm2, %%mm2 \n\t"\ - "punpcklbw %%mm2, %%mm2 \n\t"\ - "punpcklbw %%mm2, %%mm2 \n\t"\ - "psubusb %%mm4, %%mm2 \n\t"\ - "movq %%mm2, %%mm3 \n\t"\ - "psubusb %%mm4, %%mm3 \n\t"\ - "psubb %%mm3, %%mm2 \n\t"\ - "movq %1, %%mm3 \n\t"\ - "movq %2, %%mm4 \n\t"\ - "pxor %%mm6, %%mm3 \n\t"\ - "pxor %%mm6, %%mm4 \n\t"\ - "paddusb %%mm2, %%mm3 \n\t"\ - "psubusb %%mm2, %%mm4 \n\t"\ - "pxor %%mm6, %%mm3 \n\t"\ - "pxor %%mm6, %%mm4 \n\t"\ - "paddusb %%mm2, %%mm2 \n\t"\ - "packsswb %%mm1, %%mm0 \n\t"\ - "pcmpgtb %%mm0, %%mm7 \n\t"\ - "pxor %%mm7, %%mm0 \n\t"\ - "psubb %%mm7, %%mm0 \n\t"\ - "movq %%mm0, %%mm1 \n\t"\ - "psubusb %%mm2, %%mm0 \n\t"\ - "psubb %%mm0, %%mm1 \n\t"\ - "pand %5, %%mm1 \n\t"\ - "psrlw $2, %%mm1 \n\t"\ - "pxor %%mm7, %%mm1 \n\t"\ - "psubb %%mm7, %%mm1 \n\t"\ - "movq %0, %%mm5 \n\t"\ - "movq %3, %%mm6 \n\t"\ - "psubb %%mm1, %%mm5 \n\t"\ - "paddb %%mm1, %%mm6 \n\t" + "pxor %%mm7, %%mm7 \n\t"\ + "movq %0, %%mm0 \n\t"\ + "movq %0, %%mm1 \n\t"\ + "movq %3, %%mm2 \n\t"\ + "movq %3, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "psubw %%mm2, %%mm0 \n\t"\ + "psubw %%mm3, %%mm1 \n\t"\ + "movq %1, %%mm2 \n\t"\ + "movq %1, %%mm3 \n\t"\ + "movq %2, %%mm4 \n\t"\ + "movq %2, %%mm5 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ + "punpckhbw %%mm7, %%mm5 \n\t"\ + "psubw %%mm2, %%mm4 \n\t"\ + "psubw %%mm3, %%mm5 \n\t"\ + "psllw $2, %%mm4 \n\t"\ + "psllw $2, %%mm5 \n\t"\ + "paddw %%mm0, %%mm4 \n\t"\ + "paddw %%mm1, %%mm5 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "pcmpgtw %%mm4, %%mm6 \n\t"\ + "pcmpgtw %%mm5, %%mm7 \n\t"\ + "pxor %%mm6, %%mm4 \n\t"\ + "pxor %%mm7, %%mm5 \n\t"\ + "psubw %%mm6, %%mm4 \n\t"\ + "psubw %%mm7, %%mm5 \n\t"\ + "psrlw $3, %%mm4 \n\t"\ + "psrlw $3, %%mm5 \n\t"\ + "packuswb %%mm5, %%mm4 \n\t"\ + "packsswb %%mm7, %%mm6 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movd %4, %%mm2 \n\t"\ + "punpcklbw %%mm2, %%mm2 \n\t"\ + "punpcklbw %%mm2, %%mm2 \n\t"\ + "punpcklbw %%mm2, %%mm2 \n\t"\ + "psubusb %%mm4, %%mm2 \n\t"\ + "movq %%mm2, %%mm3 \n\t"\ + "psubusb %%mm4, %%mm3 \n\t"\ + "psubb %%mm3, %%mm2 \n\t"\ + "movq %1, %%mm3 \n\t"\ + "movq %2, %%mm4 \n\t"\ + "pxor %%mm6, %%mm3 \n\t"\ + "pxor %%mm6, %%mm4 \n\t"\ + "paddusb %%mm2, %%mm3 \n\t"\ + "psubusb %%mm2, %%mm4 \n\t"\ + "pxor %%mm6, %%mm3 \n\t"\ + "pxor %%mm6, %%mm4 \n\t"\ + "paddusb %%mm2, %%mm2 \n\t"\ + "packsswb %%mm1, %%mm0 \n\t"\ + "pcmpgtb %%mm0, %%mm7 \n\t"\ + "pxor %%mm7, %%mm0 \n\t"\ + "psubb %%mm7, %%mm0 \n\t"\ + "movq %%mm0, %%mm1 \n\t"\ + "psubusb %%mm2, %%mm0 \n\t"\ + "psubb %%mm0, %%mm1 \n\t"\ + "pand %5, %%mm1 \n\t"\ + "psrlw $2, %%mm1 \n\t"\ + "pxor %%mm7, %%mm1 \n\t"\ + "psubb %%mm7, %%mm1 \n\t"\ + "movq %0, %%mm5 \n\t"\ + "movq %3, %%mm6 \n\t"\ + "psubb %%mm1, %%mm5 \n\t"\ + "paddb %%mm1, %%mm6 \n\t" static void h263_v_loop_filter_mmx(uint8_t *src, int stride, int qscale){ const int strength= ff_h263_loop_filter_strength[qscale]; @@ -605,10 +605,10 @@ H263_LOOP_FILTER - "movq %%mm3, %1 \n\t" - "movq %%mm4, %2 \n\t" - "movq %%mm5, %0 \n\t" - "movq %%mm6, %3 \n\t" + "movq %%mm3, %1 \n\t" + "movq %%mm4, %2 \n\t" + "movq %%mm5, %0 \n\t" + "movq %%mm6, %3 \n\t" : "+m" (*(uint64_t*)(src - 2*stride)), "+m" (*(uint64_t*)(src - 1*stride)), "+m" (*(uint64_t*)(src + 0*stride)), @@ -619,21 +619,21 @@ static inline void transpose4x4(uint8_t *dst, uint8_t *src, int dst_stride, int src_stride){ asm volatile( //FIXME could save 1 instruction if done as 8x4 ... - "movd %4, %%mm0 \n\t" - "movd %5, %%mm1 \n\t" - "movd %6, %%mm2 \n\t" - "movd %7, %%mm3 \n\t" - "punpcklbw %%mm1, %%mm0 \n\t" - "punpcklbw %%mm3, %%mm2 \n\t" - "movq %%mm0, %%mm1 \n\t" - "punpcklwd %%mm2, %%mm0 \n\t" - "punpckhwd %%mm2, %%mm1 \n\t" - "movd %%mm0, %0 \n\t" - "punpckhdq %%mm0, %%mm0 \n\t" - "movd %%mm0, %1 \n\t" - "movd %%mm1, %2 \n\t" - "punpckhdq %%mm1, %%mm1 \n\t" - "movd %%mm1, %3 \n\t" + "movd %4, %%mm0 \n\t" + "movd %5, %%mm1 \n\t" + "movd %6, %%mm2 \n\t" + "movd %7, %%mm3 \n\t" + "punpcklbw %%mm1, %%mm0 \n\t" + "punpcklbw %%mm3, %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "punpcklwd %%mm2, %%mm0 \n\t" + "punpckhwd %%mm2, %%mm1 \n\t" + "movd %%mm0, %0 \n\t" + "punpckhdq %%mm0, %%mm0 \n\t" + "movd %%mm0, %1 \n\t" + "movd %%mm1, %2 \n\t" + "punpckhdq %%mm1, %%mm1 \n\t" + "movd %%mm1, %3 \n\t" : "=m" (*(uint32_t*)(dst + 0*dst_stride)), "=m" (*(uint32_t*)(dst + 1*dst_stride)), @@ -666,30 +666,30 @@ ); asm volatile( - "movq %%mm5, %%mm1 \n\t" - "movq %%mm4, %%mm0 \n\t" - "punpcklbw %%mm3, %%mm5 \n\t" - "punpcklbw %%mm6, %%mm4 \n\t" - "punpckhbw %%mm3, %%mm1 \n\t" - "punpckhbw %%mm6, %%mm0 \n\t" - "movq %%mm5, %%mm3 \n\t" - "movq %%mm1, %%mm6 \n\t" - "punpcklwd %%mm4, %%mm5 \n\t" - "punpcklwd %%mm0, %%mm1 \n\t" - "punpckhwd %%mm4, %%mm3 \n\t" - "punpckhwd %%mm0, %%mm6 \n\t" - "movd %%mm5, (%0) \n\t" - "punpckhdq %%mm5, %%mm5 \n\t" - "movd %%mm5, (%0,%2) \n\t" - "movd %%mm3, (%0,%2,2) \n\t" - "punpckhdq %%mm3, %%mm3 \n\t" - "movd %%mm3, (%0,%3) \n\t" - "movd %%mm1, (%1) \n\t" - "punpckhdq %%mm1, %%mm1 \n\t" - "movd %%mm1, (%1,%2) \n\t" - "movd %%mm6, (%1,%2,2) \n\t" - "punpckhdq %%mm6, %%mm6 \n\t" - "movd %%mm6, (%1,%3) \n\t" + "movq %%mm5, %%mm1 \n\t" + "movq %%mm4, %%mm0 \n\t" + "punpcklbw %%mm3, %%mm5 \n\t" + "punpcklbw %%mm6, %%mm4 \n\t" + "punpckhbw %%mm3, %%mm1 \n\t" + "punpckhbw %%mm6, %%mm0 \n\t" + "movq %%mm5, %%mm3 \n\t" + "movq %%mm1, %%mm6 \n\t" + "punpcklwd %%mm4, %%mm5 \n\t" + "punpcklwd %%mm0, %%mm1 \n\t" + "punpckhwd %%mm4, %%mm3 \n\t" + "punpckhwd %%mm0, %%mm6 \n\t" + "movd %%mm5, (%0) \n\t" + "punpckhdq %%mm5, %%mm5 \n\t" + "movd %%mm5, (%0,%2) \n\t" + "movd %%mm3, (%0,%2,2) \n\t" + "punpckhdq %%mm3, %%mm3 \n\t" + "movd %%mm3, (%0,%3) \n\t" + "movd %%mm1, (%1) \n\t" + "punpckhdq %%mm1, %%mm1 \n\t" + "movd %%mm1, (%1,%2) \n\t" + "movd %%mm6, (%1,%2,2) \n\t" + "punpckhdq %%mm6, %%mm6 \n\t" + "movd %%mm6, (%1,%3) \n\t" :: "r" (src), "r" (src + 4*stride), "r" ((long) stride ), @@ -705,26 +705,26 @@ "pxor %%mm0,%%mm0\n" "pxor %%mm7,%%mm7\n" "1:\n" - "movq (%0),%%mm2\n" /* mm2 = pix[0-7] */ - "movq 8(%0),%%mm3\n" /* mm3 = pix[8-15] */ + "movq (%0),%%mm2\n" /* mm2 = pix[0-7] */ + "movq 8(%0),%%mm3\n" /* mm3 = pix[8-15] */ - "movq %%mm2,%%mm1\n" /* mm1 = mm2 = pix[0-7] */ + "movq %%mm2,%%mm1\n" /* mm1 = mm2 = pix[0-7] */ - "punpckhbw %%mm0,%%mm1\n" /* mm1 = [pix4-7] */ - "punpcklbw %%mm0,%%mm2\n" /* mm2 = [pix0-3] */ + "punpckhbw %%mm0,%%mm1\n" /* mm1 = [pix4-7] */ + "punpcklbw %%mm0,%%mm2\n" /* mm2 = [pix0-3] */ - "movq %%mm3,%%mm4\n" /* mm4 = mm3 = pix[8-15] */ - "punpckhbw %%mm0,%%mm3\n" /* mm3 = [pix12-15] */ - "punpcklbw %%mm0,%%mm4\n" /* mm4 = [pix8-11] */ + "movq %%mm3,%%mm4\n" /* mm4 = mm3 = pix[8-15] */ + "punpckhbw %%mm0,%%mm3\n" /* mm3 = [pix12-15] */ + "punpcklbw %%mm0,%%mm4\n" /* mm4 = [pix8-11] */ - "pmaddwd %%mm1,%%mm1\n" /* mm1 = (pix0^2+pix1^2,pix2^2+pix3^2) */ - "pmaddwd %%mm2,%%mm2\n" /* mm2 = (pix4^2+pix5^2,pix6^2+pix7^2) */ + "pmaddwd %%mm1,%%mm1\n" /* mm1 = (pix0^2+pix1^2,pix2^2+pix3^2) */ + "pmaddwd %%mm2,%%mm2\n" /* mm2 = (pix4^2+pix5^2,pix6^2+pix7^2) */ "pmaddwd %%mm3,%%mm3\n" "pmaddwd %%mm4,%%mm4\n" - "paddd %%mm1,%%mm2\n" /* mm2 = (pix0^2+pix1^2+pix4^2+pix5^2, - pix2^2+pix3^2+pix6^2+pix7^2) */ + "paddd %%mm1,%%mm2\n" /* mm2 = (pix0^2+pix1^2+pix4^2+pix5^2, + pix2^2+pix3^2+pix6^2+pix7^2) */ "paddd %%mm3,%%mm4\n" "paddd %%mm2,%%mm7\n" @@ -734,7 +734,7 @@ "jnz 1b\n" "movq %%mm7,%%mm1\n" - "psrlq $32, %%mm7\n" /* shift hi dword to lo */ + "psrlq $32, %%mm7\n" /* shift hi dword to lo */ "paddd %%mm7,%%mm1\n" "movd %%mm1,%1\n" : "+r" (pix), "=r"(tmp) : "r" ((long)line_size) : "%ecx" ); @@ -746,13 +746,13 @@ asm volatile ( "movl %4,%%ecx\n" "shr $1,%%ecx\n" - "pxor %%mm0,%%mm0\n" /* mm0 = 0 */ - "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */ + "pxor %%mm0,%%mm0\n" /* mm0 = 0 */ + "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */ "1:\n" - "movq (%0),%%mm1\n" /* mm1 = pix1[0][0-7] */ - "movq (%1),%%mm2\n" /* mm2 = pix2[0][0-7] */ - "movq (%0,%3),%%mm3\n" /* mm3 = pix1[1][0-7] */ - "movq (%1,%3),%%mm4\n" /* mm4 = pix2[1][0-7] */ + "movq (%0),%%mm1\n" /* mm1 = pix1[0][0-7] */ + "movq (%1),%%mm2\n" /* mm2 = pix2[0][0-7] */ + "movq (%0,%3),%%mm3\n" /* mm3 = pix1[1][0-7] */ + "movq (%1,%3),%%mm4\n" /* mm4 = pix2[1][0-7] */ /* todo: mm1-mm2, mm3-mm4 */ /* algo: substract mm1 from mm2 with saturation and vice versa */ @@ -773,16 +773,16 @@ "punpckhbw %%mm0,%%mm2\n" "punpckhbw %%mm0,%%mm4\n" - "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */ - "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */ + "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */ + "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */ "pmaddwd %%mm2,%%mm2\n" "pmaddwd %%mm4,%%mm4\n" "pmaddwd %%mm1,%%mm1\n" "pmaddwd %%mm3,%%mm3\n" - "lea (%0,%3,2), %0\n" /* pix1 += 2*line_size */ - "lea (%1,%3,2), %1\n" /* pix2 += 2*line_size */ + "lea (%0,%3,2), %0\n" /* pix1 += 2*line_size */ + "lea (%1,%3,2), %1\n" /* pix2 += 2*line_size */ "paddd %%mm2,%%mm1\n" "paddd %%mm4,%%mm3\n" @@ -793,7 +793,7 @@ "jnz 1b\n" "movq %%mm7,%%mm1\n" - "psrlq $32, %%mm7\n" /* shift hi dword to lo */ + "psrlq $32, %%mm7\n" /* shift hi dword to lo */ "paddd %%mm7,%%mm1\n" "movd %%mm1,%2\n" : "+r" (pix1), "+r" (pix2), "=r"(tmp) @@ -806,13 +806,13 @@ int tmp; asm volatile ( "movl %4,%%ecx\n" - "pxor %%mm0,%%mm0\n" /* mm0 = 0 */ - "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */ + "pxor %%mm0,%%mm0\n" /* mm0 = 0 */ + "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */ "1:\n" - "movq (%0),%%mm1\n" /* mm1 = pix1[0-7] */ - "movq (%1),%%mm2\n" /* mm2 = pix2[0-7] */ - "movq 8(%0),%%mm3\n" /* mm3 = pix1[8-15] */ - "movq 8(%1),%%mm4\n" /* mm4 = pix2[8-15] */ + "movq (%0),%%mm1\n" /* mm1 = pix1[0-7] */ + "movq (%1),%%mm2\n" /* mm2 = pix2[0-7] */ + "movq 8(%0),%%mm3\n" /* mm3 = pix1[8-15] */ + "movq 8(%1),%%mm4\n" /* mm4 = pix2[8-15] */ /* todo: mm1-mm2, mm3-mm4 */ /* algo: substract mm1 from mm2 with saturation and vice versa */ @@ -833,8 +833,8 @@ "punpckhbw %%mm0,%%mm2\n" "punpckhbw %%mm0,%%mm4\n" - "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */ - "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */ + "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */ + "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */ "pmaddwd %%mm2,%%mm2\n" "pmaddwd %%mm4,%%mm4\n" @@ -853,7 +853,7 @@ "jnz 1b\n" "movq %%mm7,%%mm1\n" - "psrlq $32, %%mm7\n" /* shift hi dword to lo */ + "psrlq $32, %%mm7\n" /* shift hi dword to lo */ "paddd %%mm7,%%mm1\n" "movd %%mm1,%2\n" : "+r" (pix1), "+r" (pix2), "=r"(tmp) @@ -866,13 +866,13 @@ int tmp; asm volatile ( "shr $1,%2\n" - "pxor %%xmm0,%%xmm0\n" /* mm0 = 0 */ - "pxor %%xmm7,%%xmm7\n" /* mm7 holds the sum */ + "pxor %%xmm0,%%xmm0\n" /* mm0 = 0 */ + "pxor %%xmm7,%%xmm7\n" /* mm7 holds the sum */ "1:\n" - "movdqu (%0),%%xmm1\n" /* mm1 = pix1[0][0-15] */ - "movdqu (%1),%%xmm2\n" /* mm2 = pix2[0][0-15] */ - "movdqu (%0,%4),%%xmm3\n" /* mm3 = pix1[1][0-15] */ - "movdqu (%1,%4),%%xmm4\n" /* mm4 = pix2[1][0-15] */ + "movdqu (%0),%%xmm1\n" /* mm1 = pix1[0][0-15] */ + "movdqu (%1),%%xmm2\n" /* mm2 = pix2[0][0-15] */ + "movdqu (%0,%4),%%xmm3\n" /* mm3 = pix1[1][0-15] */ + "movdqu (%1,%4),%%xmm4\n" /* mm4 = pix2[1][0-15] */ /* todo: mm1-mm2, mm3-mm4 */ /* algo: substract mm1 from mm2 with saturation and vice versa */ @@ -893,16 +893,16 @@ "punpckhbw %%xmm0,%%xmm2\n" "punpckhbw %%xmm0,%%xmm4\n" - "punpcklbw %%xmm0,%%xmm1\n" /* mm1 now spread over (mm1,mm2) */ - "punpcklbw %%xmm0,%%xmm3\n" /* mm4 now spread over (mm3,mm4) */ + "punpcklbw %%xmm0,%%xmm1\n" /* mm1 now spread over (mm1,mm2) */ + "punpcklbw %%xmm0,%%xmm3\n" /* mm4 now spread over (mm3,mm4) */ "pmaddwd %%xmm2,%%xmm2\n" "pmaddwd %%xmm4,%%xmm4\n" "pmaddwd %%xmm1,%%xmm1\n" "pmaddwd %%xmm3,%%xmm3\n" - "lea (%0,%4,2), %0\n" /* pix1 += 2*line_size */ - "lea (%1,%4,2), %1\n" /* pix2 += 2*line_size */ + "lea (%0,%4,2), %0\n" /* pix1 += 2*line_size */ + "lea (%1,%4,2), %1\n" /* pix2 += 2*line_size */ "paddd %%xmm2,%%xmm1\n" "paddd %%xmm4,%%xmm3\n" @@ -913,10 +913,10 @@ "jnz 1b\n" "movdqa %%xmm7,%%xmm1\n" - "psrldq $8, %%xmm7\n" /* shift hi qword to lo */ + "psrldq $8, %%xmm7\n" /* shift hi qword to lo */ "paddd %%xmm1,%%xmm7\n" "movdqa %%xmm7,%%xmm1\n" - "psrldq $4, %%xmm7\n" /* shift hi dword to lo */ + "psrldq $4, %%xmm7\n" /* shift hi dword to lo */ "paddd %%xmm1,%%xmm7\n" "movd %%xmm7,%3\n" : "+r" (pix1), "+r" (pix2), "+r"(h), "=r"(tmp) @@ -1427,18 +1427,18 @@ static void diff_bytes_mmx(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w){ long i=0; asm volatile( - "1: \n\t" - "movq (%2, %0), %%mm0 \n\t" - "movq (%1, %0), %%mm1 \n\t" - "psubb %%mm0, %%mm1 \n\t" - "movq %%mm1, (%3, %0) \n\t" - "movq 8(%2, %0), %%mm0 \n\t" - "movq 8(%1, %0), %%mm1 \n\t" - "psubb %%mm0, %%mm1 \n\t" - "movq %%mm1, 8(%3, %0) \n\t" - "add $16, %0 \n\t" - "cmp %4, %0 \n\t" - " jb 1b \n\t" + "1: \n\t" + "movq (%2, %0), %%mm0 \n\t" + "movq (%1, %0), %%mm1 \n\t" + "psubb %%mm0, %%mm1 \n\t" + "movq %%mm1, (%3, %0) \n\t" + "movq 8(%2, %0), %%mm0 \n\t" + "movq 8(%1, %0), %%mm1 \n\t" + "psubb %%mm0, %%mm1 \n\t" + "movq %%mm1, 8(%3, %0) \n\t" + "add $16, %0 \n\t" + "cmp %4, %0 \n\t" + " jb 1b \n\t" : "+r" (i) : "r"(src1), "r"(src2), "r"(dst), "r"((long)w-15) ); @@ -1451,24 +1451,24 @@ uint8_t l, lt; asm volatile( - "1: \n\t" - "movq -1(%1, %0), %%mm0 \n\t" // LT - "movq (%1, %0), %%mm1 \n\t" // T - "movq -1(%2, %0), %%mm2 \n\t" // L - "movq (%2, %0), %%mm3 \n\t" // X - "movq %%mm2, %%mm4 \n\t" // L - "psubb %%mm0, %%mm2 \n\t" - "paddb %%mm1, %%mm2 \n\t" // L + T - LT - "movq %%mm4, %%mm5 \n\t" // L - "pmaxub %%mm1, %%mm4 \n\t" // max(T, L) - "pminub %%mm5, %%mm1 \n\t" // min(T, L) - "pminub %%mm2, %%mm4 \n\t" - "pmaxub %%mm1, %%mm4 \n\t" - "psubb %%mm4, %%mm3 \n\t" // dst - pred - "movq %%mm3, (%3, %0) \n\t" - "add $8, %0 \n\t" - "cmp %4, %0 \n\t" - " jb 1b \n\t" + "1: \n\t" + "movq -1(%1, %0), %%mm0 \n\t" // LT + "movq (%1, %0), %%mm1 \n\t" // T + "movq -1(%2, %0), %%mm2 \n\t" // L + "movq (%2, %0), %%mm3 \n\t" // X + "movq %%mm2, %%mm4 \n\t" // L + "psubb %%mm0, %%mm2 \n\t" + "paddb %%mm1, %%mm2 \n\t" // L + T - LT + "movq %%mm4, %%mm5 \n\t" // L + "pmaxub %%mm1, %%mm4 \n\t" // max(T, L) + "pminub %%mm5, %%mm1 \n\t" // min(T, L) + "pminub %%mm2, %%mm4 \n\t" + "pmaxub %%mm1, %%mm4 \n\t" + "psubb %%mm4, %%mm3 \n\t" // dst - pred + "movq %%mm3, (%3, %0) \n\t" + "add $8, %0 \n\t" + "cmp %4, %0 \n\t" + " jb 1b \n\t" : "+r" (i) : "r"(src1), "r"(src2), "r"(dst), "r"((long)w) ); @@ -1483,12 +1483,12 @@ } #define LBUTTERFLY2(a1,b1,a2,b2)\ - "paddw " #b1 ", " #a1 " \n\t"\ - "paddw " #b2 ", " #a2 " \n\t"\ - "paddw " #b1 ", " #b1 " \n\t"\ - "paddw " #b2 ", " #b2 " \n\t"\ - "psubw " #a1 ", " #b1 " \n\t"\ - "psubw " #a2 ", " #b2 " \n\t" + "paddw " #b1 ", " #a1 " \n\t"\ + "paddw " #b2 ", " #a2 " \n\t"\ + "paddw " #b1 ", " #b1 " \n\t"\ + "paddw " #b2 ", " #b2 " \n\t"\ + "psubw " #a1 ", " #b1 " \n\t"\ + "psubw " #a2 ", " #b2 " \n\t" #define HADAMARD48\ LBUTTERFLY2(%%mm0, %%mm1, %%mm2, %%mm3)\ @@ -1499,33 +1499,33 @@ LBUTTERFLY2(%%mm2, %%mm6, %%mm3, %%mm7)\ #define MMABS(a,z)\ - "pxor " #z ", " #z " \n\t"\ - "pcmpgtw " #a ", " #z " \n\t"\ - "pxor " #z ", " #a " \n\t"\ - "psubw " #z ", " #a " \n\t" + "pxor " #z ", " #z " \n\t"\ + "pcmpgtw " #a ", " #z " \n\t"\ + "pxor " #z ", " #a " \n\t"\ + "psubw " #z ", " #a " \n\t" #define MMABS_SUM(a,z, sum)\ - "pxor " #z ", " #z " \n\t"\ - "pcmpgtw " #a ", " #z " \n\t"\ - "pxor " #z ", " #a " \n\t"\ - "psubw " #z ", " #a " \n\t"\ - "paddusw " #a ", " #sum " \n\t" + "pxor " #z ", " #z " \n\t"\ + "pcmpgtw " #a ", " #z " \n\t"\ + "pxor " #z ", " #a " \n\t"\ + "psubw " #z ", " #a " \n\t"\ + "paddusw " #a ", " #sum " \n\t" #define MMABS_MMX2(a,z)\ - "pxor " #z ", " #z " \n\t"\ - "psubw " #a ", " #z " \n\t"\ - "pmaxsw " #z ", " #a " \n\t" + "pxor " #z ", " #z " \n\t"\ + "psubw " #a ", " #z " \n\t"\ + "pmaxsw " #z ", " #a " \n\t" #define MMABS_SUM_MMX2(a,z, sum)\ - "pxor " #z ", " #z " \n\t"\ - "psubw " #a ", " #z " \n\t"\ - "pmaxsw " #z ", " #a " \n\t"\ - "paddusw " #a ", " #sum " \n\t" + "pxor " #z ", " #z " \n\t"\ + "psubw " #a ", " #z " \n\t"\ + "pmaxsw " #z ", " #a " \n\t"\ + "paddusw " #a ", " #sum " \n\t" #define SBUTTERFLY(a,b,t,n)\ - "movq " #a ", " #t " \n\t" /* abcd */\ - "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\ - "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\ + "movq " #a ", " #t " \n\t" /* abcd */\ + "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\ + "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\ #define TRANSPOSE4(a,b,c,d,t)\ SBUTTERFLY(a,b,t,wd) /* a=aebf t=cgdh */\ @@ -1534,16 +1534,16 @@ SBUTTERFLY(t,b,c,dq) /* t=cgko c=dhlp */ #define LOAD4(o, a, b, c, d)\ - "movq "#o"(%1), " #a " \n\t"\ - "movq "#o"+16(%1), " #b " \n\t"\ - "movq "#o"+32(%1), " #c " \n\t"\ - "movq "#o"+48(%1), " #d " \n\t" + "movq "#o"(%1), " #a " \n\t"\ + "movq "#o"+16(%1), " #b " \n\t"\ + "movq "#o"+32(%1), " #c " \n\t"\ + "movq "#o"+48(%1), " #d " \n\t" #define STORE4(o, a, b, c, d)\ - "movq "#a", "#o"(%1) \n\t"\ - "movq "#b", "#o"+16(%1) \n\t"\ - "movq "#c", "#o"+32(%1) \n\t"\ - "movq "#d", "#o"+48(%1) \n\t"\ + "movq "#a", "#o"(%1) \n\t"\ + "movq "#b", "#o"+16(%1) \n\t"\ + "movq "#c", "#o"+32(%1) \n\t"\ + "movq "#d", "#o"+48(%1) \n\t"\ static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride, int h){ uint64_t temp[16] __align8; @@ -1559,12 +1559,12 @@ HADAMARD48 - "movq %%mm7, 112(%1) \n\t" + "movq %%mm7, 112(%1) \n\t" TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7) STORE4(0 , %%mm0, %%mm3, %%mm7, %%mm2) - "movq 112(%1), %%mm7 \n\t" + "movq 112(%1), %%mm7 \n\t" TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0) STORE4(64, %%mm4, %%mm7, %%mm0, %%mm6) @@ -1573,23 +1573,23 @@ HADAMARD48 - "movq %%mm7, 120(%1) \n\t" + "movq %%mm7, 120(%1) \n\t" TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7) STORE4(8 , %%mm0, %%mm3, %%mm7, %%mm2) - "movq 120(%1), %%mm7 \n\t" + "movq 120(%1), %%mm7 \n\t" TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0) - "movq %%mm7, %%mm5 \n\t"//FIXME remove - "movq %%mm6, %%mm7 \n\t" - "movq %%mm0, %%mm6 \n\t" + "movq %%mm7, %%mm5 \n\t"//FIXME remove + "movq %%mm6, %%mm7 \n\t" + "movq %%mm0, %%mm6 \n\t" // STORE4(72, %%mm4, %%mm7, %%mm0, %%mm6) //FIXME remove LOAD4(64, %%mm0, %%mm1, %%mm2, %%mm3) // LOAD4(72, %%mm4, %%mm5, %%mm6, %%mm7) HADAMARD48 - "movq %%mm7, 64(%1) \n\t" + "movq %%mm7, 64(%1) \n\t" MMABS(%%mm0, %%mm7) MMABS_SUM(%%mm1, %%mm7, %%mm0) MMABS_SUM(%%mm2, %%mm7, %%mm0) @@ -1597,15 +1597,15 @@ MMABS_SUM(%%mm4, %%mm7, %%mm0) MMABS_SUM(%%mm5, %%mm7, %%mm0) MMABS_SUM(%%mm6, %%mm7, %%mm0) - "movq 64(%1), %%mm1 \n\t" + "movq 64(%1), %%mm1 \n\t" MMABS_SUM(%%mm1, %%mm7, %%mm0) - "movq %%mm0, 64(%1) \n\t" + "movq %%mm0, 64(%1) \n\t" LOAD4(0 , %%mm0, %%mm1, %%mm2, %%mm3) LOAD4(8 , %%mm4, %%mm5, %%mm6, %%mm7) HADAMARD48 - "movq %%mm7, (%1) \n\t" + "movq %%mm7, (%1) \n\t" MMABS(%%mm0, %%mm7) MMABS_SUM(%%mm1, %%mm7, %%mm0) MMABS_SUM(%%mm2, %%mm7, %%mm0) @@ -1613,18 +1613,18 @@ MMABS_SUM(%%mm4, %%mm7, %%mm0) MMABS_SUM(%%mm5, %%mm7, %%mm0) MMABS_SUM(%%mm6, %%mm7, %%mm0) - "movq (%1), %%mm1 \n\t" + "movq (%1), %%mm1 \n\t" MMABS_SUM(%%mm1, %%mm7, %%mm0) - "movq 64(%1), %%mm1 \n\t" + "movq 64(%1), %%mm1 \n\t" MMABS_SUM(%%mm1, %%mm7, %%mm0) - "movq %%mm0, %%mm1 \n\t" - "psrlq $32, %%mm0 \n\t" - "paddusw %%mm1, %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "psrlq $16, %%mm0 \n\t" - "paddusw %%mm1, %%mm0 \n\t" - "movd %%mm0, %0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "psrlq $32, %%mm0 \n\t" + "paddusw %%mm1, %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "psrlq $16, %%mm0 \n\t" + "paddusw %%mm1, %%mm0 \n\t" + "movd %%mm0, %0 \n\t" : "=r" (sum) : "r"(temp) @@ -1646,12 +1646,12 @@ HADAMARD48 - "movq %%mm7, 112(%1) \n\t" + "movq %%mm7, 112(%1) \n\t" TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7) STORE4(0 , %%mm0, %%mm3, %%mm7, %%mm2) - "movq 112(%1), %%mm7 \n\t" + "movq 112(%1), %%mm7 \n\t" TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0) STORE4(64, %%mm4, %%mm7, %%mm0, %%mm6) @@ -1660,23 +1660,23 @@ HADAMARD48 - "movq %%mm7, 120(%1) \n\t" + "movq %%mm7, 120(%1) \n\t" TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7) STORE4(8 , %%mm0, %%mm3, %%mm7, %%mm2) - "movq 120(%1), %%mm7 \n\t" + "movq 120(%1), %%mm7 \n\t" TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0) - "movq %%mm7, %%mm5 \n\t"//FIXME remove - "movq %%mm6, %%mm7 \n\t" - "movq %%mm0, %%mm6 \n\t" + "movq %%mm7, %%mm5 \n\t"//FIXME remove + "movq %%mm6, %%mm7 \n\t" + "movq %%mm0, %%mm6 \n\t" // STORE4(72, %%mm4, %%mm7, %%mm0, %%mm6) //FIXME remove LOAD4(64, %%mm0, %%mm1, %%mm2, %%mm3) // LOAD4(72, %%mm4, %%mm5, %%mm6, %%mm7) HADAMARD48 - "movq %%mm7, 64(%1) \n\t" + "movq %%mm7, 64(%1) \n\t" MMABS_MMX2(%%mm0, %%mm7) MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm2, %%mm7, %%mm0) @@ -1684,15 +1684,15 @@ MMABS_SUM_MMX2(%%mm4, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm5, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm6, %%mm7, %%mm0) - "movq 64(%1), %%mm1 \n\t" + "movq 64(%1), %%mm1 \n\t" MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0) - "movq %%mm0, 64(%1) \n\t" + "movq %%mm0, 64(%1) \n\t" LOAD4(0 , %%mm0, %%mm1, %%mm2, %%mm3) LOAD4(8 , %%mm4, %%mm5, %%mm6, %%mm7) HADAMARD48 - "movq %%mm7, (%1) \n\t" + "movq %%mm7, (%1) \n\t" MMABS_MMX2(%%mm0, %%mm7) MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm2, %%mm7, %%mm0) @@ -1700,16 +1700,16 @@ MMABS_SUM_MMX2(%%mm4, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm5, %%mm7, %%mm0) MMABS_SUM_MMX2(%%mm6, %%mm7, %%mm0) - "movq (%1), %%mm1 \n\t" + "movq (%1), %%mm1 \n\t" MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0) - "movq 64(%1), %%mm1 \n\t" + "movq 64(%1), %%mm1 \n\t" MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0) "pshufw $0x0E, %%mm0, %%mm1 \n\t" - "paddusw %%mm1, %%mm0 \n\t" + "paddusw %%mm1, %%mm0 \n\t" "pshufw $0x01, %%mm0, %%mm1 \n\t" - "paddusw %%mm1, %%mm0 \n\t" - "movd %%mm0, %0 \n\t" + "paddusw %%mm1, %%mm0 \n\t" + "movd %%mm0, %0 \n\t" : "=r" (sum) : "r"(temp) @@ -1726,24 +1726,24 @@ #define put_no_rnd_pixels16_mmx(a,b,c,d) put_pixels16_mmx(a,b,c,d) #define QPEL_V_LOW(m3,m4,m5,m6, pw_20, pw_3, rnd, in0, in1, in2, in7, out, OP)\ - "paddw " #m4 ", " #m3 " \n\t" /* x1 */\ - "movq "MANGLE(ff_pw_20)", %%mm4 \n\t" /* 20 */\ - "pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\ - "movq "#in7", " #m3 " \n\t" /* d */\ - "movq "#in0", %%mm5 \n\t" /* D */\ - "paddw " #m3 ", %%mm5 \n\t" /* x4 */\ - "psubw %%mm5, %%mm4 \n\t" /* 20x1 - x4 */\ - "movq "#in1", %%mm5 \n\t" /* C */\ - "movq "#in2", %%mm6 \n\t" /* B */\ - "paddw " #m6 ", %%mm5 \n\t" /* x3 */\ - "paddw " #m5 ", %%mm6 \n\t" /* x2 */\ - "paddw %%mm6, %%mm6 \n\t" /* 2x2 */\ - "psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\ - "pmullw "MANGLE(ff_pw_3)", %%mm5 \n\t" /* -6x2 + 3x3 */\ - "paddw " #rnd ", %%mm4 \n\t" /* x2 */\ - "paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\ - "psraw $5, %%mm5 \n\t"\ - "packuswb %%mm5, %%mm5 \n\t"\ + "paddw " #m4 ", " #m3 " \n\t" /* x1 */\ + "movq "MANGLE(ff_pw_20)", %%mm4 \n\t" /* 20 */\ + "pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\ + "movq "#in7", " #m3 " \n\t" /* d */\ + "movq "#in0", %%mm5 \n\t" /* D */\ + "paddw " #m3 ", %%mm5 \n\t" /* x4 */\ + "psubw %%mm5, %%mm4 \n\t" /* 20x1 - x4 */\ + "movq "#in1", %%mm5 \n\t" /* C */\ + "movq "#in2", %%mm6 \n\t" /* B */\ + "paddw " #m6 ", %%mm5 \n\t" /* x3 */\ + "paddw " #m5 ", %%mm6 \n\t" /* x2 */\ + "paddw %%mm6, %%mm6 \n\t" /* 2x2 */\ + "psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\ + "pmullw "MANGLE(ff_pw_3)", %%mm5 \n\t" /* -6x2 + 3x3 */\ + "paddw " #rnd ", %%mm4 \n\t" /* x2 */\ + "paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\ + "psraw $5, %%mm5 \n\t"\ + "packuswb %%mm5, %%mm5 \n\t"\ OP(%%mm5, out, %%mm7, d) #define QPEL_BASE(OPNAME, ROUNDER, RND, OP_MMX2, OP_3DNOW)\ @@ -1751,116 +1751,116 @@ uint64_t temp;\ \ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\ - "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\ - "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\ - "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\ - "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\ - "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\ - "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\ - "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\ - "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\ - "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\ - "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\ - "psllq $24, %%mm4 \n\t" /* 000ABCDE */\ - "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\ - "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\ - "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\ - "paddw %%mm3, %%mm5 \n\t" /* b */\ - "paddw %%mm2, %%mm6 \n\t" /* c */\ - "paddw %%mm5, %%mm5 \n\t" /* 2b */\ - "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\ - "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\ - "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\ - "paddw %%mm4, %%mm0 \n\t" /* a */\ - "paddw %%mm1, %%mm5 \n\t" /* d */\ - "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\ - "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\ - "paddw %6, %%mm6 \n\t"\ - "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ - "psraw $5, %%mm0 \n\t"\ - "movq %%mm0, %5 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\ + "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\ + "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\ + "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\ + "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\ + "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\ + "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\ + "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\ + "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\ + "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\ + "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\ + "psllq $24, %%mm4 \n\t" /* 000ABCDE */\ + "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\ + "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\ + "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\ + "paddw %%mm3, %%mm5 \n\t" /* b */\ + "paddw %%mm2, %%mm6 \n\t" /* c */\ + "paddw %%mm5, %%mm5 \n\t" /* 2b */\ + "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\ + "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\ + "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\ + "paddw %%mm4, %%mm0 \n\t" /* a */\ + "paddw %%mm1, %%mm5 \n\t" /* d */\ + "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\ + "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\ + "paddw %6, %%mm6 \n\t"\ + "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ + "psraw $5, %%mm0 \n\t"\ + "movq %%mm0, %5 \n\t"\ /* mm1=EFGH, mm2=DEFG, mm3=CDEF, mm4=BCDE, mm7=0 */\ \ - "movq 5(%0), %%mm0 \n\t" /* FGHIJKLM */\ - "movq %%mm0, %%mm5 \n\t" /* FGHIJKLM */\ - "movq %%mm0, %%mm6 \n\t" /* FGHIJKLM */\ - "psrlq $8, %%mm0 \n\t" /* GHIJKLM0 */\ - "psrlq $16, %%mm5 \n\t" /* HIJKLM00 */\ - "punpcklbw %%mm7, %%mm0 \n\t" /* 0G0H0I0J */\ - "punpcklbw %%mm7, %%mm5 \n\t" /* 0H0I0J0K */\ - "paddw %%mm0, %%mm2 \n\t" /* b */\ - "paddw %%mm5, %%mm3 \n\t" /* c */\ - "paddw %%mm2, %%mm2 \n\t" /* 2b */\ - "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\ - "movq %%mm6, %%mm2 \n\t" /* FGHIJKLM */\ - "psrlq $24, %%mm6 \n\t" /* IJKLM000 */\ - "punpcklbw %%mm7, %%mm2 \n\t" /* 0F0G0H0I */\ - "punpcklbw %%mm7, %%mm6 \n\t" /* 0I0J0K0L */\ - "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\ - "paddw %%mm2, %%mm1 \n\t" /* a */\ - "paddw %%mm6, %%mm4 \n\t" /* d */\ - "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\ - "psubw %%mm4, %%mm3 \n\t" /* - 6b +3c - d */\ - "paddw %6, %%mm1 \n\t"\ - "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b +3c - d */\ - "psraw $5, %%mm3 \n\t"\ - "movq %5, %%mm1 \n\t"\ - "packuswb %%mm3, %%mm1 \n\t"\ + "movq 5(%0), %%mm0 \n\t" /* FGHIJKLM */\ + "movq %%mm0, %%mm5 \n\t" /* FGHIJKLM */\ + "movq %%mm0, %%mm6 \n\t" /* FGHIJKLM */\ + "psrlq $8, %%mm0 \n\t" /* GHIJKLM0 */\ + "psrlq $16, %%mm5 \n\t" /* HIJKLM00 */\ + "punpcklbw %%mm7, %%mm0 \n\t" /* 0G0H0I0J */\ + "punpcklbw %%mm7, %%mm5 \n\t" /* 0H0I0J0K */\ + "paddw %%mm0, %%mm2 \n\t" /* b */\ + "paddw %%mm5, %%mm3 \n\t" /* c */\ + "paddw %%mm2, %%mm2 \n\t" /* 2b */\ + "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\ + "movq %%mm6, %%mm2 \n\t" /* FGHIJKLM */\ + "psrlq $24, %%mm6 \n\t" /* IJKLM000 */\ + "punpcklbw %%mm7, %%mm2 \n\t" /* 0F0G0H0I */\ + "punpcklbw %%mm7, %%mm6 \n\t" /* 0I0J0K0L */\ + "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\ + "paddw %%mm2, %%mm1 \n\t" /* a */\ + "paddw %%mm6, %%mm4 \n\t" /* d */\ + "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\ + "psubw %%mm4, %%mm3 \n\t" /* - 6b +3c - d */\ + "paddw %6, %%mm1 \n\t"\ + "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b +3c - d */\ + "psraw $5, %%mm3 \n\t"\ + "movq %5, %%mm1 \n\t"\ + "packuswb %%mm3, %%mm1 \n\t"\ OP_MMX2(%%mm1, (%1),%%mm4, q)\ /* mm0= GHIJ, mm2=FGHI, mm5=HIJK, mm6=IJKL, mm7=0 */\ \ - "movq 9(%0), %%mm1 \n\t" /* JKLMNOPQ */\ - "movq %%mm1, %%mm4 \n\t" /* JKLMNOPQ */\ - "movq %%mm1, %%mm3 \n\t" /* JKLMNOPQ */\ - "psrlq $8, %%mm1 \n\t" /* KLMNOPQ0 */\ - "psrlq $16, %%mm4 \n\t" /* LMNOPQ00 */\ - "punpcklbw %%mm7, %%mm1 \n\t" /* 0K0L0M0N */\ - "punpcklbw %%mm7, %%mm4 \n\t" /* 0L0M0N0O */\ - "paddw %%mm1, %%mm5 \n\t" /* b */\ - "paddw %%mm4, %%mm0 \n\t" /* c */\ - "paddw %%mm5, %%mm5 \n\t" /* 2b */\ - "psubw %%mm5, %%mm0 \n\t" /* c - 2b */\ - "movq %%mm3, %%mm5 \n\t" /* JKLMNOPQ */\ - "psrlq $24, %%mm3 \n\t" /* MNOPQ000 */\ - "pmullw "MANGLE(ff_pw_3)", %%mm0 \n\t" /* 3c - 6b */\ - "punpcklbw %%mm7, %%mm3 \n\t" /* 0M0N0O0P */\ - "paddw %%mm3, %%mm2 \n\t" /* d */\ - "psubw %%mm2, %%mm0 \n\t" /* -6b + 3c - d */\ - "movq %%mm5, %%mm2 \n\t" /* JKLMNOPQ */\ - "punpcklbw %%mm7, %%mm2 \n\t" /* 0J0K0L0M */\ - "punpckhbw %%mm7, %%mm5 \n\t" /* 0N0O0P0Q */\ - "paddw %%mm2, %%mm6 \n\t" /* a */\ - "pmullw "MANGLE(ff_pw_20)", %%mm6 \n\t" /* 20a */\ - "paddw %6, %%mm0 \n\t"\ - "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ - "psraw $5, %%mm0 \n\t"\ + "movq 9(%0), %%mm1 \n\t" /* JKLMNOPQ */\ + "movq %%mm1, %%mm4 \n\t" /* JKLMNOPQ */\ + "movq %%mm1, %%mm3 \n\t" /* JKLMNOPQ */\ + "psrlq $8, %%mm1 \n\t" /* KLMNOPQ0 */\ + "psrlq $16, %%mm4 \n\t" /* LMNOPQ00 */\ + "punpcklbw %%mm7, %%mm1 \n\t" /* 0K0L0M0N */\ + "punpcklbw %%mm7, %%mm4 \n\t" /* 0L0M0N0O */\ + "paddw %%mm1, %%mm5 \n\t" /* b */\ + "paddw %%mm4, %%mm0 \n\t" /* c */\ + "paddw %%mm5, %%mm5 \n\t" /* 2b */\ + "psubw %%mm5, %%mm0 \n\t" /* c - 2b */\ + "movq %%mm3, %%mm5 \n\t" /* JKLMNOPQ */\ + "psrlq $24, %%mm3 \n\t" /* MNOPQ000 */\ + "pmullw "MANGLE(ff_pw_3)", %%mm0 \n\t" /* 3c - 6b */\ + "punpcklbw %%mm7, %%mm3 \n\t" /* 0M0N0O0P */\ + "paddw %%mm3, %%mm2 \n\t" /* d */\ + "psubw %%mm2, %%mm0 \n\t" /* -6b + 3c - d */\ + "movq %%mm5, %%mm2 \n\t" /* JKLMNOPQ */\ + "punpcklbw %%mm7, %%mm2 \n\t" /* 0J0K0L0M */\ + "punpckhbw %%mm7, %%mm5 \n\t" /* 0N0O0P0Q */\ + "paddw %%mm2, %%mm6 \n\t" /* a */\ + "pmullw "MANGLE(ff_pw_20)", %%mm6 \n\t" /* 20a */\ + "paddw %6, %%mm0 \n\t"\ + "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ + "psraw $5, %%mm0 \n\t"\ /* mm1=KLMN, mm2=JKLM, mm3=MNOP, mm4=LMNO, mm5=NOPQ mm7=0 */\ \ - "paddw %%mm5, %%mm3 \n\t" /* a */\ - "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0O0P0Q0Q */\ - "paddw %%mm4, %%mm6 \n\t" /* b */\ - "pshufw $0xBE, %%mm5, %%mm4 \n\t" /* 0P0Q0Q0P */\ - "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0Q0Q0P0O */\ - "paddw %%mm1, %%mm4 \n\t" /* c */\ - "paddw %%mm2, %%mm5 \n\t" /* d */\ - "paddw %%mm6, %%mm6 \n\t" /* 2b */\ - "psubw %%mm6, %%mm4 \n\t" /* c - 2b */\ - "pmullw "MANGLE(ff_pw_20)", %%mm3 \n\t" /* 20a */\ - "pmullw "MANGLE(ff_pw_3)", %%mm4 \n\t" /* 3c - 6b */\ - "psubw %%mm5, %%mm3 \n\t" /* -6b + 3c - d */\ - "paddw %6, %%mm4 \n\t"\ - "paddw %%mm3, %%mm4 \n\t" /* 20a - 6b + 3c - d */\ - "psraw $5, %%mm4 \n\t"\ - "packuswb %%mm4, %%mm0 \n\t"\ + "paddw %%mm5, %%mm3 \n\t" /* a */\ + "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0O0P0Q0Q */\ + "paddw %%mm4, %%mm6 \n\t" /* b */\ + "pshufw $0xBE, %%mm5, %%mm4 \n\t" /* 0P0Q0Q0P */\ + "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0Q0Q0P0O */\ + "paddw %%mm1, %%mm4 \n\t" /* c */\ + "paddw %%mm2, %%mm5 \n\t" /* d */\ + "paddw %%mm6, %%mm6 \n\t" /* 2b */\ + "psubw %%mm6, %%mm4 \n\t" /* c - 2b */\ + "pmullw "MANGLE(ff_pw_20)", %%mm3 \n\t" /* 20a */\ + "pmullw "MANGLE(ff_pw_3)", %%mm4 \n\t" /* 3c - 6b */\ + "psubw %%mm5, %%mm3 \n\t" /* -6b + 3c - d */\ + "paddw %6, %%mm4 \n\t"\ + "paddw %%mm3, %%mm4 \n\t" /* 20a - 6b + 3c - d */\ + "psraw $5, %%mm4 \n\t"\ + "packuswb %%mm4, %%mm0 \n\t"\ OP_MMX2(%%mm0, 8(%1), %%mm4, q)\ \ - "add %3, %0 \n\t"\ - "add %4, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add %3, %0 \n\t"\ + "add %4, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(src), "+c"(dst), "+m"(h)\ : "d"((long)srcStride), "S"((long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(temp), "m"(ROUNDER)\ : "memory"\ @@ -1890,21 +1890,21 @@ temp[14]= (src[14]+src[15])*20 - (src[13]+src[16])*6 + (src[12]+src[16])*3 - (src[11]+src[15]);\ temp[15]= (src[15]+src[16])*20 - (src[14]+src[16])*6 + (src[13]+src[15])*3 - (src[12]+src[14]);\ asm volatile(\ - "movq (%0), %%mm0 \n\t"\ - "movq 8(%0), %%mm1 \n\t"\ - "paddw %2, %%mm0 \n\t"\ - "paddw %2, %%mm1 \n\t"\ - "psraw $5, %%mm0 \n\t"\ - "psraw $5, %%mm1 \n\t"\ - "packuswb %%mm1, %%mm0 \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 8(%0), %%mm1 \n\t"\ + "paddw %2, %%mm0 \n\t"\ + "paddw %2, %%mm1 \n\t"\ + "psraw $5, %%mm0 \n\t"\ + "psraw $5, %%mm1 \n\t"\ + "packuswb %%mm1, %%mm0 \n\t"\ OP_3DNOW(%%mm0, (%1), %%mm1, q)\ - "movq 16(%0), %%mm0 \n\t"\ - "movq 24(%0), %%mm1 \n\t"\ - "paddw %2, %%mm0 \n\t"\ - "paddw %2, %%mm1 \n\t"\ - "psraw $5, %%mm0 \n\t"\ - "psraw $5, %%mm1 \n\t"\ - "packuswb %%mm1, %%mm0 \n\t"\ + "movq 16(%0), %%mm0 \n\t"\ + "movq 24(%0), %%mm1 \n\t"\ + "paddw %2, %%mm0 \n\t"\ + "paddw %2, %%mm1 \n\t"\ + "psraw $5, %%mm0 \n\t"\ + "psraw $5, %%mm1 \n\t"\ + "packuswb %%mm1, %%mm0 \n\t"\ OP_3DNOW(%%mm0, 8(%1), %%mm1, q)\ :: "r"(temp), "r"(dst), "m"(ROUNDER)\ : "memory"\ @@ -1918,62 +1918,62 @@ uint64_t temp;\ \ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\ - "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\ - "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\ - "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\ - "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\ - "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\ - "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\ - "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\ - "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\ - "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\ - "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\ - "psllq $24, %%mm4 \n\t" /* 000ABCDE */\ - "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\ - "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\ - "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\ - "paddw %%mm3, %%mm5 \n\t" /* b */\ - "paddw %%mm2, %%mm6 \n\t" /* c */\ - "paddw %%mm5, %%mm5 \n\t" /* 2b */\ - "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\ - "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\ - "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\ - "paddw %%mm4, %%mm0 \n\t" /* a */\ - "paddw %%mm1, %%mm5 \n\t" /* d */\ - "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\ - "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\ - "paddw %6, %%mm6 \n\t"\ - "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ - "psraw $5, %%mm0 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\ + "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\ + "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\ + "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\ + "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\ + "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\ + "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\ + "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\ + "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\ + "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\ + "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\ + "psllq $24, %%mm4 \n\t" /* 000ABCDE */\ + "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\ + "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\ + "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\ + "paddw %%mm3, %%mm5 \n\t" /* b */\ + "paddw %%mm2, %%mm6 \n\t" /* c */\ + "paddw %%mm5, %%mm5 \n\t" /* 2b */\ + "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\ + "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\ + "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\ + "paddw %%mm4, %%mm0 \n\t" /* a */\ + "paddw %%mm1, %%mm5 \n\t" /* d */\ + "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\ + "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\ + "paddw %6, %%mm6 \n\t"\ + "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\ + "psraw $5, %%mm0 \n\t"\ /* mm1=EFGH, mm2=DEFG, mm3=CDEF, mm4=BCDE, mm7=0 */\ \ - "movd 5(%0), %%mm5 \n\t" /* FGHI */\ - "punpcklbw %%mm7, %%mm5 \n\t" /* 0F0G0H0I */\ - "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0G0H0I0I */\ - "paddw %%mm5, %%mm1 \n\t" /* a */\ - "paddw %%mm6, %%mm2 \n\t" /* b */\ - "pshufw $0xBE, %%mm5, %%mm6 \n\t" /* 0H0I0I0H */\ - "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0I0I0H0G */\ - "paddw %%mm6, %%mm3 \n\t" /* c */\ - "paddw %%mm5, %%mm4 \n\t" /* d */\ - "paddw %%mm2, %%mm2 \n\t" /* 2b */\ - "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\ - "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\ - "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\ - "psubw %%mm4, %%mm3 \n\t" /* -6b + 3c - d */\ - "paddw %6, %%mm1 \n\t"\ - "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b + 3c - d */\ - "psraw $5, %%mm3 \n\t"\ - "packuswb %%mm3, %%mm0 \n\t"\ + "movd 5(%0), %%mm5 \n\t" /* FGHI */\ + "punpcklbw %%mm7, %%mm5 \n\t" /* 0F0G0H0I */\ + "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0G0H0I0I */\ + "paddw %%mm5, %%mm1 \n\t" /* a */\ + "paddw %%mm6, %%mm2 \n\t" /* b */\ + "pshufw $0xBE, %%mm5, %%mm6 \n\t" /* 0H0I0I0H */\ + "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0I0I0H0G */\ + "paddw %%mm6, %%mm3 \n\t" /* c */\ + "paddw %%mm5, %%mm4 \n\t" /* d */\ + "paddw %%mm2, %%mm2 \n\t" /* 2b */\ + "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\ + "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\ + "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\ + "psubw %%mm4, %%mm3 \n\t" /* -6b + 3c - d */\ + "paddw %6, %%mm1 \n\t"\ + "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b + 3c - d */\ + "psraw $5, %%mm3 \n\t"\ + "packuswb %%mm3, %%mm0 \n\t"\ OP_MMX2(%%mm0, (%1), %%mm4, q)\ \ - "add %3, %0 \n\t"\ - "add %4, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add %3, %0 \n\t"\ + "add %4, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(src), "+c"(dst), "+m"(h)\ : "S"((long)srcStride), "D"((long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(temp), "m"(ROUNDER)\ : "memory"\ @@ -1995,13 +1995,13 @@ temp[ 6]= (src[ 6]+src[ 7])*20 - (src[ 5]+src[ 8])*6 + (src[ 4]+src[ 8])*3 - (src[ 3]+src[ 7]);\ temp[ 7]= (src[ 7]+src[ 8])*20 - (src[ 6]+src[ 8])*6 + (src[ 5]+src[ 7])*3 - (src[ 4]+src[ 6]);\ asm volatile(\ - "movq (%0), %%mm0 \n\t"\ - "movq 8(%0), %%mm1 \n\t"\ - "paddw %2, %%mm0 \n\t"\ - "paddw %2, %%mm1 \n\t"\ - "psraw $5, %%mm0 \n\t"\ - "psraw $5, %%mm1 \n\t"\ - "packuswb %%mm1, %%mm0 \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 8(%0), %%mm1 \n\t"\ + "paddw %2, %%mm0 \n\t"\ + "paddw %2, %%mm1 \n\t"\ + "psraw $5, %%mm0 \n\t"\ + "psraw $5, %%mm1 \n\t"\ + "packuswb %%mm1, %%mm0 \n\t"\ OP_3DNOW(%%mm0, (%1), %%mm1, q)\ :: "r"(temp), "r"(dst), "m"(ROUNDER)\ :"memory"\ @@ -2020,24 +2020,24 @@ \ /*FIXME unroll */\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq (%0), %%mm1 \n\t"\ - "movq 8(%0), %%mm2 \n\t"\ - "movq 8(%0), %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "movq %%mm0, (%1) \n\t"\ - "movq %%mm1, 17*8(%1) \n\t"\ - "movq %%mm2, 2*17*8(%1) \n\t"\ - "movq %%mm3, 3*17*8(%1) \n\t"\ - "add $8, %1 \n\t"\ - "add %3, %0 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq (%0), %%mm1 \n\t"\ + "movq 8(%0), %%mm2 \n\t"\ + "movq 8(%0), %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "movq %%mm0, (%1) \n\t"\ + "movq %%mm1, 17*8(%1) \n\t"\ + "movq %%mm2, 2*17*8(%1) \n\t"\ + "movq %%mm3, 3*17*8(%1) \n\t"\ + "add $8, %1 \n\t"\ + "add %3, %0 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+r" (src), "+r" (temp_ptr), "+r"(count)\ : "r" ((long)srcStride)\ : "memory"\ @@ -2048,42 +2048,42 @@ \ /*FIXME reorder for speed */\ asm volatile(\ - /*"pxor %%mm7, %%mm7 \n\t"*/\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq 8(%0), %%mm1 \n\t"\ - "movq 16(%0), %%mm2 \n\t"\ - "movq 24(%0), %%mm3 \n\t"\ + /*"pxor %%mm7, %%mm7 \n\t"*/\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 8(%0), %%mm1 \n\t"\ + "movq 16(%0), %%mm2 \n\t"\ + "movq 24(%0), %%mm3 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 16(%0), 8(%0), (%0), 32(%0), (%1), OP)\ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 8(%0), (%0), (%0), 40(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, (%0), (%0), 8(%0), 48(%0), (%1), OP)\ \ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, (%0), 8(%0), 16(%0), 56(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 8(%0), 16(%0), 24(%0), 64(%0), (%1), OP)\ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 16(%0), 24(%0), 32(%0), 72(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 24(%0), 32(%0), 40(%0), 80(%0), (%1), OP)\ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 32(%0), 40(%0), 48(%0), 88(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 40(%0), 48(%0), 56(%0), 96(%0), (%1), OP)\ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 48(%0), 56(%0), 64(%0),104(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 56(%0), 64(%0), 72(%0),112(%0), (%1), OP)\ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 64(%0), 72(%0), 80(%0),120(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 72(%0), 80(%0), 88(%0),128(%0), (%1), OP)\ \ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 80(%0), 88(%0), 96(%0),128(%0), (%1, %3), OP)\ - "add %4, %1 \n\t" \ + "add %4, %1 \n\t" \ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 88(%0), 96(%0),104(%0),120(%0), (%1), OP)\ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 96(%0),104(%0),112(%0),112(%0), (%1, %3), OP)\ \ - "add $136, %0 \n\t"\ - "add %6, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add $136, %0 \n\t"\ + "add %6, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ \ : "+r"(temp_ptr), "+r"(dst), "+g"(count)\ : "r"((long)dstStride), "r"(2*(long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(ROUNDER), "g"(4-14*(long)dstStride)\ @@ -2098,18 +2098,18 @@ \ /*FIXME unroll */\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq (%0), %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm1 \n\t"\ - "movq %%mm0, (%1) \n\t"\ - "movq %%mm1, 9*8(%1) \n\t"\ - "add $8, %1 \n\t"\ - "add %3, %0 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq (%0), %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm1 \n\t"\ + "movq %%mm0, (%1) \n\t"\ + "movq %%mm1, 9*8(%1) \n\t"\ + "add $8, %1 \n\t"\ + "add %3, %0 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+r" (src), "+r" (temp_ptr), "+r"(count)\ : "r" ((long)srcStride)\ : "memory"\ @@ -2120,30 +2120,30 @@ \ /*FIXME reorder for speed */\ asm volatile(\ - /*"pxor %%mm7, %%mm7 \n\t"*/\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq 8(%0), %%mm1 \n\t"\ - "movq 16(%0), %%mm2 \n\t"\ - "movq 24(%0), %%mm3 \n\t"\ + /*"pxor %%mm7, %%mm7 \n\t"*/\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 8(%0), %%mm1 \n\t"\ + "movq 16(%0), %%mm2 \n\t"\ + "movq 24(%0), %%mm3 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 16(%0), 8(%0), (%0), 32(%0), (%1), OP)\ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 8(%0), (%0), (%0), 40(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, (%0), (%0), 8(%0), 48(%0), (%1), OP)\ \ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, (%0), 8(%0), 16(%0), 56(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 8(%0), 16(%0), 24(%0), 64(%0), (%1), OP)\ \ QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 16(%0), 24(%0), 32(%0), 64(%0), (%1, %3), OP)\ - "add %4, %1 \n\t"\ + "add %4, %1 \n\t"\ QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 24(%0), 32(%0), 40(%0), 56(%0), (%1), OP)\ QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 32(%0), 40(%0), 48(%0), 48(%0), (%1, %3), OP)\ \ - "add $72, %0 \n\t"\ - "add %6, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add $72, %0 \n\t"\ + "add %6, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ \ : "+r"(temp_ptr), "+r"(dst), "+g"(count)\ : "r"((long)dstStride), "r"(2*(long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(ROUNDER), "g"(4-6*(long)dstStride)\ @@ -2374,15 +2374,15 @@ OPNAME ## mpeg4_qpel16_v_lowpass_ ## MMX(dst, halfH, stride, 16);\ } -#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t" +#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t" #define AVG_3DNOW_OP(a,b,temp, size) \ -"mov" #size " " #b ", " #temp " \n\t"\ -"pavgusb " #temp ", " #a " \n\t"\ -"mov" #size " " #a ", " #b " \n\t" +"mov" #size " " #b ", " #temp " \n\t"\ +"pavgusb " #temp ", " #a " \n\t"\ +"mov" #size " " #a ", " #b " \n\t" #define AVG_MMX2_OP(a,b,temp, size) \ -"mov" #size " " #b ", " #temp " \n\t"\ -"pavgb " #temp ", " #a " \n\t"\ -"mov" #size " " #a ", " #b " \n\t" +"mov" #size " " #b ", " #temp " \n\t"\ +"pavgb " #temp ", " #a " \n\t"\ +"mov" #size " " #a ", " #b " \n\t" QPEL_BASE(put_ , ff_pw_16, _ , PUT_OP, PUT_OP) QPEL_BASE(avg_ , ff_pw_16, _ , AVG_MMX2_OP, AVG_3DNOW_OP) @@ -2410,40 +2410,40 @@ scale<<= 16 + 1 - BASIS_SHIFT + RECON_SHIFT; asm volatile( - "pcmpeqw %%mm6, %%mm6 \n\t" // -1w - "psrlw $15, %%mm6 \n\t" // 1w - "pxor %%mm7, %%mm7 \n\t" - "movd %4, %%mm5 \n\t" - "punpcklwd %%mm5, %%mm5 \n\t" - "punpcklwd %%mm5, %%mm5 \n\t" - "1: \n\t" - "movq (%1, %0), %%mm0 \n\t" - "movq 8(%1, %0), %%mm1 \n\t" - "pmulhw %%mm5, %%mm0 \n\t" - "pmulhw %%mm5, %%mm1 \n\t" - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm6, %%mm1 \n\t" - "psraw $1, %%mm0 \n\t" - "psraw $1, %%mm1 \n\t" - "paddw (%2, %0), %%mm0 \n\t" - "paddw 8(%2, %0), %%mm1 \n\t" - "psraw $6, %%mm0 \n\t" - "psraw $6, %%mm1 \n\t" - "pmullw (%3, %0), %%mm0 \n\t" - "pmullw 8(%3, %0), %%mm1 \n\t" - "pmaddwd %%mm0, %%mm0 \n\t" - "pmaddwd %%mm1, %%mm1 \n\t" - "paddd %%mm1, %%mm0 \n\t" - "psrld $4, %%mm0 \n\t" - "paddd %%mm0, %%mm7 \n\t" - "add $16, %0 \n\t" - "cmp $128, %0 \n\t" //FIXME optimize & bench - " jb 1b \n\t" - "movq %%mm7, %%mm6 \n\t" - "psrlq $32, %%mm7 \n\t" - "paddd %%mm6, %%mm7 \n\t" - "psrld $2, %%mm7 \n\t" - "movd %%mm7, %0 \n\t" + "pcmpeqw %%mm6, %%mm6 \n\t" // -1w + "psrlw $15, %%mm6 \n\t" // 1w + "pxor %%mm7, %%mm7 \n\t" + "movd %4, %%mm5 \n\t" + "punpcklwd %%mm5, %%mm5 \n\t" + "punpcklwd %%mm5, %%mm5 \n\t" + "1: \n\t" + "movq (%1, %0), %%mm0 \n\t" + "movq 8(%1, %0), %%mm1 \n\t" + "pmulhw %%mm5, %%mm0 \n\t" + "pmulhw %%mm5, %%mm1 \n\t" + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm6, %%mm1 \n\t" + "psraw $1, %%mm0 \n\t" + "psraw $1, %%mm1 \n\t" + "paddw (%2, %0), %%mm0 \n\t" + "paddw 8(%2, %0), %%mm1 \n\t" + "psraw $6, %%mm0 \n\t" + "psraw $6, %%mm1 \n\t" + "pmullw (%3, %0), %%mm0 \n\t" + "pmullw 8(%3, %0), %%mm1 \n\t" + "pmaddwd %%mm0, %%mm0 \n\t" + "pmaddwd %%mm1, %%mm1 \n\t" + "paddd %%mm1, %%mm0 \n\t" + "psrld $4, %%mm0 \n\t" + "paddd %%mm0, %%mm7 \n\t" + "add $16, %0 \n\t" + "cmp $128, %0 \n\t" //FIXME optimize & bench + " jb 1b \n\t" + "movq %%mm7, %%mm6 \n\t" + "psrlq $32, %%mm7 \n\t" + "paddd %%mm6, %%mm7 \n\t" + "psrld $2, %%mm7 \n\t" + "movd %%mm7, %0 \n\t" : "+r" (i) : "r"(basis), "r"(rem), "r"(weight), "g"(scale) @@ -2457,27 +2457,27 @@ if(ABS(scale) < 256){ scale<<= 16 + 1 - BASIS_SHIFT + RECON_SHIFT; asm volatile( - "pcmpeqw %%mm6, %%mm6 \n\t" // -1w - "psrlw $15, %%mm6 \n\t" // 1w - "movd %3, %%mm5 \n\t" - "punpcklwd %%mm5, %%mm5 \n\t" - "punpcklwd %%mm5, %%mm5 \n\t" - "1: \n\t" - "movq (%1, %0), %%mm0 \n\t" - "movq 8(%1, %0), %%mm1 \n\t" - "pmulhw %%mm5, %%mm0 \n\t" - "pmulhw %%mm5, %%mm1 \n\t" - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm6, %%mm1 \n\t" - "psraw $1, %%mm0 \n\t" - "psraw $1, %%mm1 \n\t" - "paddw (%2, %0), %%mm0 \n\t" - "paddw 8(%2, %0), %%mm1 \n\t" - "movq %%mm0, (%2, %0) \n\t" - "movq %%mm1, 8(%2, %0) \n\t" - "add $16, %0 \n\t" - "cmp $128, %0 \n\t" //FIXME optimize & bench - " jb 1b \n\t" + "pcmpeqw %%mm6, %%mm6 \n\t" // -1w + "psrlw $15, %%mm6 \n\t" // 1w + "movd %3, %%mm5 \n\t" + "punpcklwd %%mm5, %%mm5 \n\t" + "punpcklwd %%mm5, %%mm5 \n\t" + "1: \n\t" + "movq (%1, %0), %%mm0 \n\t" + "movq 8(%1, %0), %%mm1 \n\t" + "pmulhw %%mm5, %%mm0 \n\t" + "pmulhw %%mm5, %%mm1 \n\t" + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm6, %%mm1 \n\t" + "psraw $1, %%mm0 \n\t" + "psraw $1, %%mm1 \n\t" + "paddw (%2, %0), %%mm0 \n\t" + "paddw 8(%2, %0), %%mm1 \n\t" + "movq %%mm0, (%2, %0) \n\t" + "movq %%mm1, 8(%2, %0) \n\t" + "add $16, %0 \n\t" + "cmp $128, %0 \n\t" //FIXME optimize & bench + " jb 1b \n\t" : "+r" (i) : "r"(basis), "r"(rem), "g"(scale) @@ -2569,10 +2569,10 @@ mm_flags = mm_support(); if (avctx->dsp_mask) { - if (avctx->dsp_mask & FF_MM_FORCE) - mm_flags |= (avctx->dsp_mask & 0xffff); - else - mm_flags &= ~(avctx->dsp_mask & 0xffff); + if (avctx->dsp_mask & FF_MM_FORCE) + mm_flags |= (avctx->dsp_mask & 0xffff); + else + mm_flags &= ~(avctx->dsp_mask & 0xffff); } #if 0 @@ -2598,7 +2598,7 @@ if(dct_algo==FF_DCT_AUTO || dct_algo==FF_DCT_MMX){ if(mm_flags & MM_SSE2){ c->fdct = ff_fdct_sse2; - }else if(mm_flags & MM_MMXEXT){ + }else if(mm_flags & MM_MMXEXT){ c->fdct = ff_fdct_mmx2; }else{ c->fdct = ff_fdct_mmx; @@ -2709,13 +2709,13 @@ c->hadamard8_diff[0]= hadamard8_diff16_mmx; c->hadamard8_diff[1]= hadamard8_diff_mmx; - c->pix_norm1 = pix_norm1_mmx; - c->sse[0] = (mm_flags & MM_SSE2) ? sse16_sse2 : sse16_mmx; - c->sse[1] = sse8_mmx; + c->pix_norm1 = pix_norm1_mmx; + c->sse[0] = (mm_flags & MM_SSE2) ? sse16_sse2 : sse16_mmx; + c->sse[1] = sse8_mmx; c->vsad[4]= vsad_intra16_mmx; - c->nsse[0] = nsse16_mmx; - c->nsse[1] = nsse8_mmx; + c->nsse[0] = nsse16_mmx; + c->nsse[1] = nsse8_mmx; if(!(avctx->flags & CODEC_FLAG_BITEXACT)){ c->vsad[0] = vsad16_mmx; } @@ -2729,7 +2729,7 @@ c->h263_v_loop_filter= h263_v_loop_filter_mmx; c->h263_h_loop_filter= h263_h_loop_filter_mmx; - c->put_h264_chroma_pixels_tab[0]= put_h264_chroma_mc8_mmx; + c->put_h264_chroma_pixels_tab[0]= put_h264_chroma_mc8_mmx; c->put_h264_chroma_pixels_tab[1]= put_h264_chroma_mc4_mmx; if (mm_flags & MM_MMXEXT) { @@ -2829,7 +2829,7 @@ dspfunc(avg_h264_qpel, 2, 4); #undef dspfunc - c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_mmx2; + c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_mmx2; c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_mmx2; c->h264_v_loop_filter_luma= h264_v_loop_filter_luma_mmx2; c->h264_h_loop_filter_luma= h264_h_loop_filter_luma_mmx2; @@ -2941,7 +2941,7 @@ dspfunc(avg_h264_qpel, 1, 8); dspfunc(avg_h264_qpel, 2, 4); - c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_3dnow; + c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_3dnow; c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_3dnow; } }
--- a/i386/dsputil_mmx_avg.h Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/dsputil_mmx_avg.h Thu Dec 22 01:10:11 2005 +0000 @@ -28,522 +28,522 @@ static void DEF(put_pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - "add %%"REG_a", %1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + "add %%"REG_a", %1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(put_pixels4_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movd (%1), %%mm0 \n\t" - "movd (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $4, %2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movd (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movd (%1), %%mm1 \n\t" - "movd (%2), %%mm2 \n\t" - "movd 4(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movd %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "movd (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movd (%1), %%mm1 \n\t" - "movd 8(%2), %%mm2 \n\t" - "movd 12(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movd %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "add $16, %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movd (%1), %%mm0 \n\t" + "movd (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $4, %2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movd (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movd (%1), %%mm1 \n\t" + "movd (%2), %%mm2 \n\t" + "movd 4(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movd %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "movd (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movd (%1), %%mm1 \n\t" + "movd 8(%2), %%mm2 \n\t" + "movd 12(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movd %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "add $16, %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); } static void DEF(put_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $8, %2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" 16(%2), %%mm0 \n\t" - PAVGB" 24(%2), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $8, %2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" 16(%2), %%mm0 \n\t" + PAVGB" 24(%2), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } static void DEF(put_no_rnd_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "pcmpeqb %%mm6, %%mm6 \n\t" - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $8, %2 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq (%2), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm6, %%mm3 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq 16(%2), %%mm2 \n\t" - "movq 24(%2), %%mm3 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm6, %%mm3 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "pcmpeqb %%mm6, %%mm6 \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $8, %2 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq (%2), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm6, %%mm3 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq 16(%2), %%mm2 \n\t" + "movq 24(%2), %%mm3 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm6, %%mm3 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } static void DEF(avg_pixels4_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movd (%1), %%mm0 \n\t" - "movd (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $4, %2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movd (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movd (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 4(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - PAVGB" (%3), %%mm1 \n\t" - "movd %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "movd (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movd (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" 8(%2), %%mm0 \n\t" - PAVGB" 12(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movd %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - PAVGB" (%3), %%mm1 \n\t" - "movd %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "add $16, %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movd (%1), %%mm0 \n\t" + "movd (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $4, %2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movd (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movd (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 4(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + PAVGB" (%3), %%mm1 \n\t" + "movd %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "movd (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movd (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" 8(%2), %%mm0 \n\t" + PAVGB" 12(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movd %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + PAVGB" (%3), %%mm1 \n\t" + "movd %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "add $16, %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); } static void DEF(avg_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $8, %2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - PAVGB" (%3), %%mm1 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" 16(%2), %%mm0 \n\t" - PAVGB" 24(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - "movq %%mm0, (%3) \n\t" - "add %5, %3 \n\t" - PAVGB" (%3), %%mm1 \n\t" - "movq %%mm1, (%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $8, %2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + PAVGB" (%3), %%mm1 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" 16(%2), %%mm0 \n\t" + PAVGB" 24(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + "movq %%mm0, (%3) \n\t" + "add %5, %3 \n\t" + PAVGB" (%3), %%mm1 \n\t" + "movq %%mm1, (%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } static void DEF(put_pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq 8(%1), %%mm2 \n\t" - "movq 8(%1, %3), %%mm3 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - PAVGB" 9(%1), %%mm2 \n\t" - PAVGB" 9(%1, %3), %%mm3 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "movq %%mm2, 8(%2) \n\t" - "movq %%mm3, 8(%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq 8(%1), %%mm2 \n\t" - "movq 8(%1, %3), %%mm3 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - PAVGB" 9(%1), %%mm2 \n\t" - PAVGB" 9(%1, %3), %%mm3 \n\t" - "add %%"REG_a", %1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "movq %%mm2, 8(%2) \n\t" - "movq %%mm3, 8(%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq 8(%1), %%mm2 \n\t" + "movq 8(%1, %3), %%mm3 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + PAVGB" 9(%1), %%mm2 \n\t" + PAVGB" 9(%1, %3), %%mm3 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "movq %%mm2, 8(%2) \n\t" + "movq %%mm3, 8(%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq 8(%1), %%mm2 \n\t" + "movq 8(%1, %3), %%mm3 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + PAVGB" 9(%1), %%mm2 \n\t" + PAVGB" 9(%1, %3), %%mm3 \n\t" + "add %%"REG_a", %1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "movq %%mm2, 8(%2) \n\t" + "movq %%mm3, 8(%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(put_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $16, %2 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" 16(%2), %%mm0 \n\t" - PAVGB" 24(%2), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $16, %2 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" 16(%2), %%mm0 \n\t" + PAVGB" 24(%2), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $2, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } static void DEF(avg_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $16, %2 \n\t" - PAVGB" (%3), %%mm0 \n\t" - PAVGB" 8(%3), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" 8(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - PAVGB" 8(%3), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - PAVGB" 16(%2), %%mm0 \n\t" - PAVGB" 24(%2), %%mm1 \n\t" - PAVGB" (%3), %%mm0 \n\t" - PAVGB" 8(%3), %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $16, %2 \n\t" + PAVGB" (%3), %%mm0 \n\t" + PAVGB" 8(%3), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" 8(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + PAVGB" 8(%3), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + PAVGB" 16(%2), %%mm0 \n\t" + PAVGB" 24(%2), %%mm1 \n\t" + PAVGB" (%3), %%mm0 \n\t" + PAVGB" 8(%3), %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $2, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } static void DEF(put_no_rnd_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { __asm __volatile( - "pcmpeqb %%mm6, %%mm6\n\t" - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "movq (%2), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm6, %%mm3 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "add %4, %1 \n\t" - "add $16, %2 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq (%2), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm6, %%mm3 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq 16(%2), %%mm2 \n\t" - "movq 24(%2), %%mm3 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm6, %%mm3 \n\t" - PAVGB" %%mm2, %%mm0 \n\t" - PAVGB" %%mm3, %%mm1 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" + "pcmpeqb %%mm6, %%mm6 \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "movq (%2), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm6, %%mm3 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "add %4, %1 \n\t" + "add $16, %2 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq (%2), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm6, %%mm3 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq 16(%2), %%mm2 \n\t" + "movq 24(%2), %%mm3 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm6, %%mm3 \n\t" + PAVGB" %%mm2, %%mm0 \n\t" + PAVGB" %%mm3, %%mm1 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $2, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); //the following should be used, though better not with gcc ... -/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) - :"r"(src1Stride), "r"(dstStride) - :"memory");*/ +/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst) + :"r"(src1Stride), "r"(dstStride) + :"memory");*/ } /* GL: this function does incorrect rounding if overflow */ @@ -551,67 +551,67 @@ { MOVQ_BONE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - "add %%"REG_a", %1 \n\t" - "psubusb %%mm6, %%mm0 \n\t" - "psubusb %%mm6, %%mm2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm3, %%mm2 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm2, (%2, %3) \n\t" - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - "psubusb %%mm6, %%mm0 \n\t" - "psubusb %%mm6, %%mm2 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm3, %%mm2 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm2, (%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + "add %%"REG_a", %1 \n\t" + "psubusb %%mm6, %%mm0 \n\t" + "psubusb %%mm6, %%mm2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm3, %%mm2 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm2, (%2, %3) \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + "psubusb %%mm6, %%mm0 \n\t" + "psubusb %%mm6, %%mm2 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm3, %%mm2 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm2, (%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(put_pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - "sub %3, %2 \n\t" - "1: \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm2, %%mm1 \n\t" - "movq %%mm0, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" %%mm1, %%mm2 \n\t" - PAVGB" %%mm0, %%mm1 \n\t" - "movq %%mm2, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D" (block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + "sub %3, %2 \n\t" + "1: \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm2, %%mm1 \n\t" + "movq %%mm0, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" %%mm1, %%mm2 \n\t" + PAVGB" %%mm0, %%mm1 \n\t" + "movq %%mm2, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D" (block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } /* GL: this function does incorrect rounding if overflow */ @@ -619,131 +619,131 @@ { MOVQ_BONE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - "sub %3, %2 \n\t" - "1: \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - "add %%"REG_a", %1 \n\t" - "psubusb %%mm6, %%mm1 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm2, %%mm1 \n\t" - "movq %%mm0, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - "psubusb %%mm6, %%mm1 \n\t" - PAVGB" %%mm1, %%mm2 \n\t" - PAVGB" %%mm0, %%mm1 \n\t" - "movq %%mm2, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D" (block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + "sub %3, %2 \n\t" + "1: \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + "add %%"REG_a", %1 \n\t" + "psubusb %%mm6, %%mm1 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm2, %%mm1 \n\t" + "movq %%mm0, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + "psubusb %%mm6, %%mm1 \n\t" + PAVGB" %%mm1, %%mm2 \n\t" + PAVGB" %%mm0, %%mm1 \n\t" + "movq %%mm2, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D" (block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(avg_pixels8)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "1: \n\t" - "movq (%2), %%mm0 \n\t" - "movq (%2, %3), %%mm1 \n\t" - PAVGB" (%1), %%mm0 \n\t" - PAVGB" (%1, %3), %%mm1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%2), %%mm0 \n\t" - "movq (%2, %3), %%mm1 \n\t" - PAVGB" (%1), %%mm0 \n\t" - PAVGB" (%1, %3), %%mm1 \n\t" - "add %%"REG_a", %1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "1: \n\t" + "movq (%2), %%mm0 \n\t" + "movq (%2, %3), %%mm1 \n\t" + PAVGB" (%1), %%mm0 \n\t" + PAVGB" (%1, %3), %%mm1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%2), %%mm0 \n\t" + "movq (%2, %3), %%mm1 \n\t" + PAVGB" (%1), %%mm0 \n\t" + PAVGB" (%1, %3), %%mm1 \n\t" + "add %%"REG_a", %1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(avg_pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm2 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm2 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" (%2, %3), %%mm2 \n\t" - "add %%"REG_a", %1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm2, (%2, %3) \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%1, %3), %%mm2 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm2 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" (%2, %3), %%mm2 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm2, (%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm2 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm2 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" (%2, %3), %%mm2 \n\t" + "add %%"REG_a", %1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm2, (%2, %3) \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%1, %3), %%mm2 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm2 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" (%2, %3), %%mm2 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm2, (%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } static void DEF(avg_pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - "sub %3, %2 \n\t" - "1: \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm2, %%mm1 \n\t" - "movq (%2, %3), %%mm3 \n\t" - "movq (%2, %%"REG_a"), %%mm4 \n\t" - PAVGB" %%mm3, %%mm0 \n\t" - PAVGB" %%mm4, %%mm1 \n\t" - "movq %%mm0, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - PAVGB" %%mm1, %%mm2 \n\t" - PAVGB" %%mm0, %%mm1 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - "movq (%2, %3), %%mm3 \n\t" - "movq (%2, %%"REG_a"), %%mm4 \n\t" - PAVGB" %%mm3, %%mm2 \n\t" - PAVGB" %%mm4, %%mm1 \n\t" - "movq %%mm2, (%2, %3) \n\t" - "movq %%mm1, (%2, %%"REG_a") \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + "sub %3, %2 \n\t" + "1: \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm2, %%mm1 \n\t" + "movq (%2, %3), %%mm3 \n\t" + "movq (%2, %%"REG_a"), %%mm4 \n\t" + PAVGB" %%mm3, %%mm0 \n\t" + PAVGB" %%mm4, %%mm1 \n\t" + "movq %%mm0, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + PAVGB" %%mm1, %%mm2 \n\t" + PAVGB" %%mm0, %%mm1 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + "movq (%2, %3), %%mm3 \n\t" + "movq (%2, %%"REG_a"), %%mm4 \n\t" + PAVGB" %%mm3, %%mm2 \n\t" + PAVGB" %%mm4, %%mm1 \n\t" + "movq %%mm2, (%2, %3) \n\t" + "movq %%mm1, (%2, %%"REG_a") \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } // Note this is not correctly rounded, but this function is only used for b frames so it doesnt matter @@ -751,41 +751,41 @@ { MOVQ_BONE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - PAVGB" 1(%1), %%mm0 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "psubusb %%mm6, %%mm2 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - PAVGB" 1(%1, %%"REG_a"), %%mm2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" %%mm1, %%mm0 \n\t" - PAVGB" %%mm2, %%mm1 \n\t" - PAVGB" (%2), %%mm0 \n\t" - PAVGB" (%2, %3), %%mm1 \n\t" - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - PAVGB" 1(%1, %3), %%mm1 \n\t" - PAVGB" 1(%1, %%"REG_a"), %%mm0 \n\t" - "add %%"REG_a", %2 \n\t" - "add %%"REG_a", %1 \n\t" - PAVGB" %%mm1, %%mm2 \n\t" - PAVGB" %%mm0, %%mm1 \n\t" - PAVGB" (%2), %%mm2 \n\t" - PAVGB" (%2, %3), %%mm1 \n\t" - "movq %%mm2, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r" ((long)line_size) - :"%"REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + PAVGB" 1(%1), %%mm0 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "psubusb %%mm6, %%mm2 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + PAVGB" 1(%1, %%"REG_a"), %%mm2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" %%mm1, %%mm0 \n\t" + PAVGB" %%mm2, %%mm1 \n\t" + PAVGB" (%2), %%mm0 \n\t" + PAVGB" (%2, %3), %%mm1 \n\t" + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + PAVGB" 1(%1, %3), %%mm1 \n\t" + PAVGB" 1(%1, %%"REG_a"), %%mm0 \n\t" + "add %%"REG_a", %2 \n\t" + "add %%"REG_a", %1 \n\t" + PAVGB" %%mm1, %%mm2 \n\t" + PAVGB" %%mm0, %%mm1 \n\t" + PAVGB" (%2), %%mm2 \n\t" + PAVGB" (%2, %3), %%mm1 \n\t" + "movq %%mm2, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r" ((long)line_size) + :"%"REG_a, "memory"); } //FIXME the following could be optimized too ...
--- a/i386/dsputil_mmx_rnd.h Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/dsputil_mmx_rnd.h Thu Dec 22 01:10:11 2005 +0000 @@ -27,206 +27,206 @@ { MOVQ_BFE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r"((long)line_size) - :REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r"((long)line_size) + :REG_a, "memory"); } static void attribute_unused DEF(put, pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { MOVQ_BFE(mm6); __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "add $8, %2 \n\t" - PAVGB(%%mm0, %%mm1, %%mm4, %%mm6) - "movq %%mm4, (%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm5, (%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 16(%2), %%mm1 \n\t" - "add %4, %1 \n\t" - "movq (%1), %%mm2 \n\t" - "movq 24(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - "add $32, %2 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%3) \n\t" - "add %5, %3 \n\t" - "movq %%mm5, (%3) \n\t" - "add %5, %3 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "add $8, %2 \n\t" + PAVGB(%%mm0, %%mm1, %%mm4, %%mm6) + "movq %%mm4, (%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm5, (%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 16(%2), %%mm1 \n\t" + "add %4, %1 \n\t" + "movq (%1), %%mm2 \n\t" + "movq 24(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + "add $32, %2 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%3) \n\t" + "add %5, %3 \n\t" + "movq %%mm5, (%3) \n\t" + "add %5, %3 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); } static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { MOVQ_BFE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "movq 8(%1), %%mm0 \n\t" - "movq 9(%1), %%mm1 \n\t" - "movq 8(%1, %3), %%mm2 \n\t" - "movq 9(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, 8(%2) \n\t" - "movq %%mm5, 8(%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm1 \n\t" - "movq (%1, %3), %%mm2 \n\t" - "movq 1(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "movq 8(%1), %%mm0 \n\t" - "movq 9(%1), %%mm1 \n\t" - "movq 8(%1, %3), %%mm2 \n\t" - "movq 9(%1, %3), %%mm3 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, 8(%2) \n\t" - "movq %%mm5, 8(%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r"((long)line_size) - :REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "movq 8(%1), %%mm0 \n\t" + "movq 9(%1), %%mm1 \n\t" + "movq 8(%1, %3), %%mm2 \n\t" + "movq 9(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, 8(%2) \n\t" + "movq %%mm5, 8(%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm1 \n\t" + "movq (%1, %3), %%mm2 \n\t" + "movq 1(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "movq 8(%1), %%mm0 \n\t" + "movq 9(%1), %%mm1 \n\t" + "movq 8(%1, %3), %%mm2 \n\t" + "movq 9(%1, %3), %%mm3 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, 8(%2) \n\t" + "movq %%mm5, 8(%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r"((long)line_size) + :REG_a, "memory"); } static void attribute_unused DEF(put, pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h) { MOVQ_BFE(mm6); __asm __volatile( - "testl $1, %0 \n\t" - " jz 1f \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "movq 8(%1), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - "add $16, %2 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%3) \n\t" - "movq %%mm5, 8(%3) \n\t" - "add %5, %3 \n\t" - "decl %0 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq (%2), %%mm1 \n\t" - "movq 8(%1), %%mm2 \n\t" - "movq 8(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%3) \n\t" - "movq %%mm5, 8(%3) \n\t" - "add %5, %3 \n\t" - "movq (%1), %%mm0 \n\t" - "movq 16(%2), %%mm1 \n\t" - "movq 8(%1), %%mm2 \n\t" - "movq 24(%2), %%mm3 \n\t" - "add %4, %1 \n\t" - PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) - "movq %%mm4, (%3) \n\t" - "movq %%mm5, 8(%3) \n\t" - "add %5, %3 \n\t" - "add $32, %2 \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" + "testl $1, %0 \n\t" + " jz 1f \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "movq 8(%1), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + "add $16, %2 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%3) \n\t" + "movq %%mm5, 8(%3) \n\t" + "add %5, %3 \n\t" + "decl %0 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq (%2), %%mm1 \n\t" + "movq 8(%1), %%mm2 \n\t" + "movq 8(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%3) \n\t" + "movq %%mm5, 8(%3) \n\t" + "add %5, %3 \n\t" + "movq (%1), %%mm0 \n\t" + "movq 16(%2), %%mm1 \n\t" + "movq 8(%1), %%mm2 \n\t" + "movq 24(%2), %%mm3 \n\t" + "add %4, %1 \n\t" + PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) + "movq %%mm4, (%3) \n\t" + "movq %%mm5, 8(%3) \n\t" + "add %5, %3 \n\t" + "add $32, %2 \n\t" + "subl $2, %0 \n\t" + "jnz 1b \n\t" #ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used - :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst) #else - :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) + :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst) #endif - :"S"((long)src1Stride), "D"((long)dstStride) - :"memory"); + :"S"((long)src1Stride), "D"((long)dstStride) + :"memory"); } static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) { MOVQ_BFE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"),%%mm2 \n\t" - PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"),%%mm0 \n\t" - PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5) - "movq %%mm4, (%2) \n\t" - "movq %%mm5, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r"((long)line_size) - :REG_a, "memory"); + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"),%%mm2 \n\t" + PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"),%%mm0 \n\t" + PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5) + "movq %%mm4, (%2) \n\t" + "movq %%mm5, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r"((long)line_size) + :REG_a, "memory"); } static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h) @@ -234,65 +234,65 @@ MOVQ_ZERO(mm7); SET_RND(mm6); // =2 for rnd and =1 for no_rnd version __asm __volatile( - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm4 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm5 \n\t" - "paddusw %%mm0, %%mm4 \n\t" - "paddusw %%mm1, %%mm5 \n\t" - "xor %%"REG_a", %%"REG_a" \n\t" - "add %3, %1 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq 1(%1, %%"REG_a"), %%mm2 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "paddusw %%mm2, %%mm0 \n\t" - "paddusw %%mm3, %%mm1 \n\t" - "paddusw %%mm6, %%mm4 \n\t" - "paddusw %%mm6, %%mm5 \n\t" - "paddusw %%mm0, %%mm4 \n\t" - "paddusw %%mm1, %%mm5 \n\t" - "psrlw $2, %%mm4 \n\t" - "psrlw $2, %%mm5 \n\t" - "packuswb %%mm5, %%mm4 \n\t" - "movq %%mm4, (%2, %%"REG_a") \n\t" - "add %3, %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm4 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddusw %%mm0, %%mm4 \n\t" + "paddusw %%mm1, %%mm5 \n\t" + "xor %%"REG_a", %%"REG_a" \n\t" + "add %3, %1 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq 1(%1, %%"REG_a"), %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "paddusw %%mm2, %%mm0 \n\t" + "paddusw %%mm3, %%mm1 \n\t" + "paddusw %%mm6, %%mm4 \n\t" + "paddusw %%mm6, %%mm5 \n\t" + "paddusw %%mm0, %%mm4 \n\t" + "paddusw %%mm1, %%mm5 \n\t" + "psrlw $2, %%mm4 \n\t" + "psrlw $2, %%mm5 \n\t" + "packuswb %%mm5, %%mm4 \n\t" + "movq %%mm4, (%2, %%"REG_a") \n\t" + "add %3, %%"REG_a" \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3 - "movq 1(%1, %%"REG_a"), %%mm4 \n\t" - "movq %%mm2, %%mm3 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "punpckhbw %%mm7, %%mm5 \n\t" - "paddusw %%mm2, %%mm4 \n\t" - "paddusw %%mm3, %%mm5 \n\t" - "paddusw %%mm6, %%mm0 \n\t" - "paddusw %%mm6, %%mm1 \n\t" - "paddusw %%mm4, %%mm0 \n\t" - "paddusw %%mm5, %%mm1 \n\t" - "psrlw $2, %%mm0 \n\t" - "psrlw $2, %%mm1 \n\t" - "packuswb %%mm1, %%mm0 \n\t" - "movq %%mm0, (%2, %%"REG_a") \n\t" - "add %3, %%"REG_a" \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3 + "movq 1(%1, %%"REG_a"), %%mm4 \n\t" + "movq %%mm2, %%mm3 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddusw %%mm2, %%mm4 \n\t" + "paddusw %%mm3, %%mm5 \n\t" + "paddusw %%mm6, %%mm0 \n\t" + "paddusw %%mm6, %%mm1 \n\t" + "paddusw %%mm4, %%mm0 \n\t" + "paddusw %%mm5, %%mm1 \n\t" + "psrlw $2, %%mm0 \n\t" + "psrlw $2, %%mm1 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "movq %%mm0, (%2, %%"REG_a") \n\t" + "add %3, %%"REG_a" \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels) - :"D"(block), "r"((long)line_size) - :REG_a, "memory"); + "subl $2, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels) + :"D"(block), "r"((long)line_size) + :REG_a, "memory"); } // avg_pixels @@ -301,16 +301,16 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movd %0, %%mm0 \n\t" - "movd %1, %%mm1 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - "movd %%mm2, %0 \n\t" - :"+m"(*block) - :"m"(*pixels) - :"memory"); - pixels += line_size; - block += line_size; + __asm __volatile( + "movd %0, %%mm0 \n\t" + "movd %1, %%mm1 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + "movd %%mm2, %0 \n\t" + :"+m"(*block) + :"m"(*pixels) + :"memory"); + pixels += line_size; + block += line_size; } while (--h); } @@ -321,16 +321,16 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %0, %%mm0 \n\t" - "movq %1, %%mm1 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - "movq %%mm2, %0 \n\t" - :"+m"(*block) - :"m"(*pixels) - :"memory"); - pixels += line_size; - block += line_size; + __asm __volatile( + "movq %0, %%mm0 \n\t" + "movq %1, %%mm1 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + "movq %%mm2, %0 \n\t" + :"+m"(*block) + :"m"(*pixels) + :"memory"); + pixels += line_size; + block += line_size; } while (--h); } @@ -340,20 +340,20 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %0, %%mm0 \n\t" - "movq %1, %%mm1 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - "movq %%mm2, %0 \n\t" - "movq 8%0, %%mm0 \n\t" - "movq 8%1, %%mm1 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - "movq %%mm2, 8%0 \n\t" - :"+m"(*block) - :"m"(*pixels) - :"memory"); - pixels += line_size; - block += line_size; + __asm __volatile( + "movq %0, %%mm0 \n\t" + "movq %1, %%mm1 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + "movq %%mm2, %0 \n\t" + "movq 8%0, %%mm0 \n\t" + "movq 8%1, %%mm1 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + "movq %%mm2, 8%0 \n\t" + :"+m"(*block) + :"m"(*pixels) + :"memory"); + pixels += line_size; + block += line_size; } while (--h); } @@ -363,18 +363,18 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %1, %%mm0 \n\t" - "movq 1%1, %%mm1 \n\t" - "movq %0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, %0 \n\t" - :"+m"(*block) - :"m"(*pixels) - :"memory"); - pixels += line_size; - block += line_size; + __asm __volatile( + "movq %1, %%mm0 \n\t" + "movq 1%1, %%mm1 \n\t" + "movq %0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, %0 \n\t" + :"+m"(*block) + :"m"(*pixels) + :"memory"); + pixels += line_size; + block += line_size; } while (--h); } @@ -383,17 +383,17 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %1, %%mm0 \n\t" - "movq %2, %%mm1 \n\t" - "movq %0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, %0 \n\t" - :"+m"(*dst) - :"m"(*src1), "m"(*src2) - :"memory"); - dst += dstStride; + __asm __volatile( + "movq %1, %%mm0 \n\t" + "movq %2, %%mm1 \n\t" + "movq %0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, %0 \n\t" + :"+m"(*dst) + :"m"(*src1), "m"(*src2) + :"memory"); + dst += dstStride; src1 += src1Stride; src2 += 8; } while (--h); @@ -404,24 +404,24 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %1, %%mm0 \n\t" - "movq 1%1, %%mm1 \n\t" - "movq %0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, %0 \n\t" - "movq 8%1, %%mm0 \n\t" - "movq 9%1, %%mm1 \n\t" - "movq 8%0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, 8%0 \n\t" - :"+m"(*block) - :"m"(*pixels) - :"memory"); - pixels += line_size; - block += line_size; + __asm __volatile( + "movq %1, %%mm0 \n\t" + "movq 1%1, %%mm1 \n\t" + "movq %0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, %0 \n\t" + "movq 8%1, %%mm0 \n\t" + "movq 9%1, %%mm1 \n\t" + "movq 8%0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, 8%0 \n\t" + :"+m"(*block) + :"m"(*pixels) + :"memory"); + pixels += line_size; + block += line_size; } while (--h); } @@ -430,23 +430,23 @@ MOVQ_BFE(mm6); JUMPALIGN(); do { - __asm __volatile( - "movq %1, %%mm0 \n\t" - "movq %2, %%mm1 \n\t" - "movq %0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, %0 \n\t" - "movq 8%1, %%mm0 \n\t" - "movq 8%2, %%mm1 \n\t" - "movq 8%0, %%mm3 \n\t" - PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) - PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) - "movq %%mm0, 8%0 \n\t" - :"+m"(*dst) - :"m"(*src1), "m"(*src2) - :"memory"); - dst += dstStride; + __asm __volatile( + "movq %1, %%mm0 \n\t" + "movq %2, %%mm1 \n\t" + "movq %0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, %0 \n\t" + "movq 8%1, %%mm0 \n\t" + "movq 8%2, %%mm1 \n\t" + "movq 8%0, %%mm3 \n\t" + PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) + PAVGB(%%mm3, %%mm2, %%mm0, %%mm6) + "movq %%mm0, 8%0 \n\t" + :"+m"(*dst) + :"m"(*src1), "m"(*src2) + :"memory"); + dst += dstStride; src1 += src1Stride; src2 += 16; } while (--h); @@ -456,39 +456,39 @@ { MOVQ_BFE(mm6); __asm __volatile( - "lea (%3, %3), %%"REG_a" \n\t" - "movq (%1), %%mm0 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5) - "movq (%2), %%mm3 \n\t" - PAVGB(%%mm3, %%mm4, %%mm0, %%mm6) - "movq (%2, %3), %%mm3 \n\t" - PAVGB(%%mm3, %%mm5, %%mm1, %%mm6) - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" + "lea (%3, %3), %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5) + "movq (%2), %%mm3 \n\t" + PAVGB(%%mm3, %%mm4, %%mm0, %%mm6) + "movq (%2, %3), %%mm3 \n\t" + PAVGB(%%mm3, %%mm5, %%mm1, %%mm6) + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" - "movq (%1, %3), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5) - "movq (%2), %%mm3 \n\t" - PAVGB(%%mm3, %%mm4, %%mm2, %%mm6) - "movq (%2, %3), %%mm3 \n\t" - PAVGB(%%mm3, %%mm5, %%mm1, %%mm6) - "movq %%mm2, (%2) \n\t" - "movq %%mm1, (%2, %3) \n\t" - "add %%"REG_a", %1 \n\t" - "add %%"REG_a", %2 \n\t" + "movq (%1, %3), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5) + "movq (%2), %%mm3 \n\t" + PAVGB(%%mm3, %%mm4, %%mm2, %%mm6) + "movq (%2, %3), %%mm3 \n\t" + PAVGB(%%mm3, %%mm5, %%mm1, %%mm6) + "movq %%mm2, (%2) \n\t" + "movq %%mm1, (%2, %3) \n\t" + "add %%"REG_a", %1 \n\t" + "add %%"REG_a", %2 \n\t" - "subl $4, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels), "+D"(block) - :"r"((long)line_size) - :REG_a, "memory"); + "subl $4, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels), "+D"(block) + :"r"((long)line_size) + :REG_a, "memory"); } // this routine is 'slightly' suboptimal but mostly unused @@ -497,73 +497,73 @@ MOVQ_ZERO(mm7); SET_RND(mm6); // =2 for rnd and =1 for no_rnd version __asm __volatile( - "movq (%1), %%mm0 \n\t" - "movq 1(%1), %%mm4 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm5 \n\t" - "paddusw %%mm0, %%mm4 \n\t" - "paddusw %%mm1, %%mm5 \n\t" - "xor %%"REG_a", %%"REG_a" \n\t" - "add %3, %1 \n\t" - ".balign 8 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq 1(%1, %%"REG_a"), %%mm2 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "paddusw %%mm2, %%mm0 \n\t" - "paddusw %%mm3, %%mm1 \n\t" - "paddusw %%mm6, %%mm4 \n\t" - "paddusw %%mm6, %%mm5 \n\t" - "paddusw %%mm0, %%mm4 \n\t" - "paddusw %%mm1, %%mm5 \n\t" - "psrlw $2, %%mm4 \n\t" - "psrlw $2, %%mm5 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "packuswb %%mm5, %%mm4 \n\t" - "pcmpeqd %%mm2, %%mm2 \n\t" - "paddb %%mm2, %%mm2 \n\t" - PAVGB(%%mm3, %%mm4, %%mm5, %%mm2) - "movq %%mm5, (%2, %%"REG_a") \n\t" - "add %3, %%"REG_a" \n\t" + "movq (%1), %%mm0 \n\t" + "movq 1(%1), %%mm4 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddusw %%mm0, %%mm4 \n\t" + "paddusw %%mm1, %%mm5 \n\t" + "xor %%"REG_a", %%"REG_a" \n\t" + "add %3, %1 \n\t" + ".balign 8 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq 1(%1, %%"REG_a"), %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "paddusw %%mm2, %%mm0 \n\t" + "paddusw %%mm3, %%mm1 \n\t" + "paddusw %%mm6, %%mm4 \n\t" + "paddusw %%mm6, %%mm5 \n\t" + "paddusw %%mm0, %%mm4 \n\t" + "paddusw %%mm1, %%mm5 \n\t" + "psrlw $2, %%mm4 \n\t" + "psrlw $2, %%mm5 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "packuswb %%mm5, %%mm4 \n\t" + "pcmpeqd %%mm2, %%mm2 \n\t" + "paddb %%mm2, %%mm2 \n\t" + PAVGB(%%mm3, %%mm4, %%mm5, %%mm2) + "movq %%mm5, (%2, %%"REG_a") \n\t" + "add %3, %%"REG_a" \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3 - "movq 1(%1, %%"REG_a"), %%mm4 \n\t" - "movq %%mm2, %%mm3 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "punpckhbw %%mm7, %%mm5 \n\t" - "paddusw %%mm2, %%mm4 \n\t" - "paddusw %%mm3, %%mm5 \n\t" - "paddusw %%mm6, %%mm0 \n\t" - "paddusw %%mm6, %%mm1 \n\t" - "paddusw %%mm4, %%mm0 \n\t" - "paddusw %%mm5, %%mm1 \n\t" - "psrlw $2, %%mm0 \n\t" - "psrlw $2, %%mm1 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "packuswb %%mm1, %%mm0 \n\t" - "pcmpeqd %%mm2, %%mm2 \n\t" - "paddb %%mm2, %%mm2 \n\t" - PAVGB(%%mm3, %%mm0, %%mm1, %%mm2) - "movq %%mm1, (%2, %%"REG_a") \n\t" - "add %3, %%"REG_a" \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3 + "movq 1(%1, %%"REG_a"), %%mm4 \n\t" + "movq %%mm2, %%mm3 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "punpckhbw %%mm7, %%mm5 \n\t" + "paddusw %%mm2, %%mm4 \n\t" + "paddusw %%mm3, %%mm5 \n\t" + "paddusw %%mm6, %%mm0 \n\t" + "paddusw %%mm6, %%mm1 \n\t" + "paddusw %%mm4, %%mm0 \n\t" + "paddusw %%mm5, %%mm1 \n\t" + "psrlw $2, %%mm0 \n\t" + "psrlw $2, %%mm1 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "pcmpeqd %%mm2, %%mm2 \n\t" + "paddb %%mm2, %%mm2 \n\t" + PAVGB(%%mm3, %%mm0, %%mm1, %%mm2) + "movq %%mm1, (%2, %%"REG_a") \n\t" + "add %3, %%"REG_a" \n\t" - "subl $2, %0 \n\t" - "jnz 1b \n\t" - :"+g"(h), "+S"(pixels) - :"D"(block), "r"((long)line_size) - :REG_a, "memory"); + "subl $2, %0 \n\t" + "jnz 1b \n\t" + :"+g"(h), "+S"(pixels) + :"D"(block), "r"((long)line_size) + :REG_a, "memory"); } //FIXME optimize
--- a/i386/fdct_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/fdct_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -30,21 +30,21 @@ // ////////////////////////////////////////////////////////////////////// -#define BITS_FRW_ACC 3 //; 2 or 3 for accuracy -#define SHIFT_FRW_COL BITS_FRW_ACC -#define SHIFT_FRW_ROW (BITS_FRW_ACC + 17 - 3) -#define RND_FRW_ROW (1 << (SHIFT_FRW_ROW-1)) -//#define RND_FRW_COL (1 << (SHIFT_FRW_COL-1)) +#define BITS_FRW_ACC 3 //; 2 or 3 for accuracy +#define SHIFT_FRW_COL BITS_FRW_ACC +#define SHIFT_FRW_ROW (BITS_FRW_ACC + 17 - 3) +#define RND_FRW_ROW (1 << (SHIFT_FRW_ROW-1)) +//#define RND_FRW_COL (1 << (SHIFT_FRW_COL-1)) //concatenated table, for forward DCT transformation static const int16_t fdct_tg_all_16[] ATTR_ALIGN(8) = { - 13036, 13036, 13036, 13036, // tg * (2<<16) + 0.5 - 27146, 27146, 27146, 27146, // tg * (2<<16) + 0.5 - -21746, -21746, -21746, -21746, // tg * (2<<16) + 0.5 + 13036, 13036, 13036, 13036, // tg * (2<<16) + 0.5 + 27146, 27146, 27146, 27146, // tg * (2<<16) + 0.5 + -21746, -21746, -21746, -21746, // tg * (2<<16) + 0.5 }; static const int16_t ocos_4_16[4] ATTR_ALIGN(8) = { - 23170, 23170, 23170, 23170, //cos * (2<<15) + 0.5 + 23170, 23170, 23170, 23170, //cos * (2<<15) + 0.5 }; static const int64_t fdct_one_corr ATTR_ALIGN(8) = 0x0001000100010001LL; @@ -351,62 +351,62 @@ static always_inline void fdct_row_sse2(const int16_t *in, int16_t *out) { asm volatile( - ".macro FDCT_ROW_SSE2_H1 i t \n\t" - "movq \\i(%0), %%xmm2 \n\t" - "movq \\i+8(%0), %%xmm0 \n\t" - "movdqa \\t+32(%1), %%xmm3 \n\t" - "movdqa \\t+48(%1), %%xmm7 \n\t" - "movdqa \\t(%1), %%xmm4 \n\t" - "movdqa \\t+16(%1), %%xmm5 \n\t" - ".endm \n\t" - ".macro FDCT_ROW_SSE2_H2 i t \n\t" - "movq \\i(%0), %%xmm2 \n\t" - "movq \\i+8(%0), %%xmm0 \n\t" - "movdqa \\t+32(%1), %%xmm3 \n\t" - "movdqa \\t+48(%1), %%xmm7 \n\t" - ".endm \n\t" - ".macro FDCT_ROW_SSE2 i \n\t" - "movq %%xmm2, %%xmm1 \n\t" - "pshuflw $27, %%xmm0, %%xmm0 \n\t" - "paddsw %%xmm0, %%xmm1 \n\t" - "psubsw %%xmm0, %%xmm2 \n\t" - "punpckldq %%xmm2, %%xmm1 \n\t" - "pshufd $78, %%xmm1, %%xmm2 \n\t" - "pmaddwd %%xmm2, %%xmm3 \n\t" - "pmaddwd %%xmm1, %%xmm7 \n\t" - "pmaddwd %%xmm5, %%xmm2 \n\t" - "pmaddwd %%xmm4, %%xmm1 \n\t" - "paddd %%xmm7, %%xmm3 \n\t" - "paddd %%xmm2, %%xmm1 \n\t" - "paddd %%xmm6, %%xmm3 \n\t" - "paddd %%xmm6, %%xmm1 \n\t" - "psrad %3, %%xmm3 \n\t" - "psrad %3, %%xmm1 \n\t" - "packssdw %%xmm3, %%xmm1 \n\t" - "movdqa %%xmm1, \\i(%4) \n\t" - ".endm \n\t" - "movdqa (%2), %%xmm6 \n\t" - "FDCT_ROW_SSE2_H1 0 0 \n\t" - "FDCT_ROW_SSE2 0 \n\t" - "FDCT_ROW_SSE2_H2 64 0 \n\t" - "FDCT_ROW_SSE2 64 \n\t" + ".macro FDCT_ROW_SSE2_H1 i t \n\t" + "movq \\i(%0), %%xmm2 \n\t" + "movq \\i+8(%0), %%xmm0 \n\t" + "movdqa \\t+32(%1), %%xmm3 \n\t" + "movdqa \\t+48(%1), %%xmm7 \n\t" + "movdqa \\t(%1), %%xmm4 \n\t" + "movdqa \\t+16(%1), %%xmm5 \n\t" + ".endm \n\t" + ".macro FDCT_ROW_SSE2_H2 i t \n\t" + "movq \\i(%0), %%xmm2 \n\t" + "movq \\i+8(%0), %%xmm0 \n\t" + "movdqa \\t+32(%1), %%xmm3 \n\t" + "movdqa \\t+48(%1), %%xmm7 \n\t" + ".endm \n\t" + ".macro FDCT_ROW_SSE2 i \n\t" + "movq %%xmm2, %%xmm1 \n\t" + "pshuflw $27, %%xmm0, %%xmm0 \n\t" + "paddsw %%xmm0, %%xmm1 \n\t" + "psubsw %%xmm0, %%xmm2 \n\t" + "punpckldq %%xmm2, %%xmm1 \n\t" + "pshufd $78, %%xmm1, %%xmm2 \n\t" + "pmaddwd %%xmm2, %%xmm3 \n\t" + "pmaddwd %%xmm1, %%xmm7 \n\t" + "pmaddwd %%xmm5, %%xmm2 \n\t" + "pmaddwd %%xmm4, %%xmm1 \n\t" + "paddd %%xmm7, %%xmm3 \n\t" + "paddd %%xmm2, %%xmm1 \n\t" + "paddd %%xmm6, %%xmm3 \n\t" + "paddd %%xmm6, %%xmm1 \n\t" + "psrad %3, %%xmm3 \n\t" + "psrad %3, %%xmm1 \n\t" + "packssdw %%xmm3, %%xmm1 \n\t" + "movdqa %%xmm1, \\i(%4) \n\t" + ".endm \n\t" + "movdqa (%2), %%xmm6 \n\t" + "FDCT_ROW_SSE2_H1 0 0 \n\t" + "FDCT_ROW_SSE2 0 \n\t" + "FDCT_ROW_SSE2_H2 64 0 \n\t" + "FDCT_ROW_SSE2 64 \n\t" - "FDCT_ROW_SSE2_H1 16 64 \n\t" - "FDCT_ROW_SSE2 16 \n\t" - "FDCT_ROW_SSE2_H2 112 64 \n\t" - "FDCT_ROW_SSE2 112 \n\t" + "FDCT_ROW_SSE2_H1 16 64 \n\t" + "FDCT_ROW_SSE2 16 \n\t" + "FDCT_ROW_SSE2_H2 112 64 \n\t" + "FDCT_ROW_SSE2 112 \n\t" - "FDCT_ROW_SSE2_H1 32 128 \n\t" - "FDCT_ROW_SSE2 32 \n\t" - "FDCT_ROW_SSE2_H2 96 128 \n\t" - "FDCT_ROW_SSE2 96 \n\t" + "FDCT_ROW_SSE2_H1 32 128 \n\t" + "FDCT_ROW_SSE2 32 \n\t" + "FDCT_ROW_SSE2_H2 96 128 \n\t" + "FDCT_ROW_SSE2 96 \n\t" - "FDCT_ROW_SSE2_H1 48 192 \n\t" - "FDCT_ROW_SSE2 48 \n\t" - "FDCT_ROW_SSE2_H2 80 192 \n\t" - "FDCT_ROW_SSE2 80 \n\t" - : - : "r" (in), "r" (tab_frw_01234567_sse2.tab_frw_01234567_sse2), "r" (fdct_r_row_sse2.fdct_r_row_sse2), "i" (SHIFT_FRW_ROW), "r" (out) + "FDCT_ROW_SSE2_H1 48 192 \n\t" + "FDCT_ROW_SSE2 48 \n\t" + "FDCT_ROW_SSE2_H2 80 192 \n\t" + "FDCT_ROW_SSE2 80 \n\t" + : + : "r" (in), "r" (tab_frw_01234567_sse2.tab_frw_01234567_sse2), "r" (fdct_r_row_sse2.fdct_r_row_sse2), "i" (SHIFT_FRW_ROW), "r" (out) ); }
--- a/i386/fft_sse.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/fft_sse.c Thu Dec 22 01:10:11 2005 +0000 @@ -45,8 +45,8 @@ void ff_fft_calc_sse(FFTContext *s, FFTComplex *z) { int ln = s->nbits; - int j, np, np2; - int nblocks, nloops; + int j, np, np2; + int nblocks, nloops; register FFTComplex *p, *q; FFTComplex *cptr, *cptr1; int k;
--- a/i386/h264dsp_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/h264dsp_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -47,9 +47,9 @@ SUMSUB_BADC( d13, s02, s13, d02 ) #define SBUTTERFLY(a,b,t,n)\ - "movq " #a ", " #t " \n\t" /* abcd */\ - "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\ - "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\ + "movq " #a ", " #t " \n\t" /* abcd */\ + "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\ + "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\ #define TRANSPOSE4(a,b,c,d,t)\ SBUTTERFLY(a,b,t,wd) /* a=aebf t=cgdh */\ @@ -369,73 +369,73 @@ /* motion compensation */ #define QPEL_H264V(A,B,C,D,E,F,OP)\ - "movd (%0), "#F" \n\t"\ - "movq "#C", %%mm6 \n\t"\ - "paddw "#D", %%mm6 \n\t"\ - "psllw $2, %%mm6 \n\t"\ - "psubw "#B", %%mm6 \n\t"\ - "psubw "#E", %%mm6 \n\t"\ - "pmullw %4, %%mm6 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, "#F" \n\t"\ - "paddw %5, "#A" \n\t"\ - "paddw "#F", "#A" \n\t"\ - "paddw "#A", %%mm6 \n\t"\ - "psraw $5, %%mm6 \n\t"\ - "packuswb %%mm6, %%mm6 \n\t"\ + "movd (%0), "#F" \n\t"\ + "movq "#C", %%mm6 \n\t"\ + "paddw "#D", %%mm6 \n\t"\ + "psllw $2, %%mm6 \n\t"\ + "psubw "#B", %%mm6 \n\t"\ + "psubw "#E", %%mm6 \n\t"\ + "pmullw %4, %%mm6 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, "#F" \n\t"\ + "paddw %5, "#A" \n\t"\ + "paddw "#F", "#A" \n\t"\ + "paddw "#A", %%mm6 \n\t"\ + "psraw $5, %%mm6 \n\t"\ + "packuswb %%mm6, %%mm6 \n\t"\ OP(%%mm6, (%1), A, d)\ - "add %3, %1 \n\t" + "add %3, %1 \n\t" #define QPEL_H264HV(A,B,C,D,E,F,OF)\ - "movd (%0), "#F" \n\t"\ - "movq "#C", %%mm6 \n\t"\ - "paddw "#D", %%mm6 \n\t"\ - "psllw $2, %%mm6 \n\t"\ - "psubw "#B", %%mm6 \n\t"\ - "psubw "#E", %%mm6 \n\t"\ - "pmullw %3, %%mm6 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, "#F" \n\t"\ - "paddw "#F", "#A" \n\t"\ - "paddw "#A", %%mm6 \n\t"\ - "movq %%mm6, "#OF"(%1) \n\t" + "movd (%0), "#F" \n\t"\ + "movq "#C", %%mm6 \n\t"\ + "paddw "#D", %%mm6 \n\t"\ + "psllw $2, %%mm6 \n\t"\ + "psubw "#B", %%mm6 \n\t"\ + "psubw "#E", %%mm6 \n\t"\ + "pmullw %3, %%mm6 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, "#F" \n\t"\ + "paddw "#F", "#A" \n\t"\ + "paddw "#A", %%mm6 \n\t"\ + "movq %%mm6, "#OF"(%1) \n\t" #define QPEL_H264(OPNAME, OP, MMX)\ static void OPNAME ## h264_qpel4_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\ int h=4;\ \ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movq %5, %%mm4 \n\t"\ - "movq %6, %%mm5 \n\t"\ - "1: \n\t"\ - "movd -1(%0), %%mm1 \n\t"\ - "movd (%0), %%mm2 \n\t"\ - "movd 1(%0), %%mm3 \n\t"\ - "movd 2(%0), %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "paddw %%mm0, %%mm1 \n\t"\ - "paddw %%mm3, %%mm2 \n\t"\ - "movd -2(%0), %%mm0 \n\t"\ - "movd 3(%0), %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "paddw %%mm3, %%mm0 \n\t"\ - "psllw $2, %%mm2 \n\t"\ - "psubw %%mm1, %%mm2 \n\t"\ - "pmullw %%mm4, %%mm2 \n\t"\ - "paddw %%mm5, %%mm0 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "psraw $5, %%mm0 \n\t"\ - "packuswb %%mm0, %%mm0 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movq %5, %%mm4 \n\t"\ + "movq %6, %%mm5 \n\t"\ + "1: \n\t"\ + "movd -1(%0), %%mm1 \n\t"\ + "movd (%0), %%mm2 \n\t"\ + "movd 1(%0), %%mm3 \n\t"\ + "movd 2(%0), %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "paddw %%mm0, %%mm1 \n\t"\ + "paddw %%mm3, %%mm2 \n\t"\ + "movd -2(%0), %%mm0 \n\t"\ + "movd 3(%0), %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "paddw %%mm3, %%mm0 \n\t"\ + "psllw $2, %%mm2 \n\t"\ + "psubw %%mm1, %%mm2 \n\t"\ + "pmullw %%mm4, %%mm2 \n\t"\ + "paddw %%mm5, %%mm0 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "psraw $5, %%mm0 \n\t"\ + "packuswb %%mm0, %%mm0 \n\t"\ OP(%%mm0, (%1),%%mm6, d)\ - "add %3, %0 \n\t"\ - "add %4, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add %3, %0 \n\t"\ + "add %4, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(src), "+c"(dst), "+m"(h)\ : "d"((long)srcStride), "S"((long)dstStride), "m"(ff_pw_5), "m"(ff_pw_16)\ : "memory"\ @@ -444,22 +444,22 @@ static void OPNAME ## h264_qpel4_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\ src -= 2*srcStride;\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movd (%0), %%mm0 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm1 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm2 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm3 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm4 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movd (%0), %%mm0 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm1 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm2 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm3 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm4 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ QPEL_H264V(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\ QPEL_H264V(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\ QPEL_H264V(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\ @@ -476,22 +476,22 @@ src -= 2*srcStride+2;\ while(w--){\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movd (%0), %%mm0 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm1 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm2 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm3 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm4 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movd (%0), %%mm0 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm1 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm2 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm3 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm4 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ QPEL_H264HV(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, 0*8*3)\ QPEL_H264HV(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, 1*8*3)\ QPEL_H264HV(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, 2*8*3)\ @@ -506,28 +506,28 @@ }\ tmp -= 3*4;\ asm volatile(\ - "movq %4, %%mm6 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "paddw 10(%0), %%mm0 \n\t"\ - "movq 2(%0), %%mm1 \n\t"\ - "paddw 8(%0), %%mm1 \n\t"\ - "movq 4(%0), %%mm2 \n\t"\ - "paddw 6(%0), %%mm2 \n\t"\ - "psubw %%mm1, %%mm0 \n\t"/*a-b (abccba)*/\ - "psraw $2, %%mm0 \n\t"/*(a-b)/4 */\ - "psubw %%mm1, %%mm0 \n\t"/*(a-b)/4-b */\ - "paddsw %%mm2, %%mm0 \n\t"\ - "psraw $2, %%mm0 \n\t"/*((a-b)/4-b)/4 */\ - "paddw %%mm6, %%mm2 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "psraw $6, %%mm0 \n\t"\ - "packuswb %%mm0, %%mm0 \n\t"\ + "movq %4, %%mm6 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "paddw 10(%0), %%mm0 \n\t"\ + "movq 2(%0), %%mm1 \n\t"\ + "paddw 8(%0), %%mm1 \n\t"\ + "movq 4(%0), %%mm2 \n\t"\ + "paddw 6(%0), %%mm2 \n\t"\ + "psubw %%mm1, %%mm0 \n\t"/*a-b (abccba)*/\ + "psraw $2, %%mm0 \n\t"/*(a-b)/4 */\ + "psubw %%mm1, %%mm0 \n\t"/*(a-b)/4-b */\ + "paddsw %%mm2, %%mm0 \n\t"\ + "psraw $2, %%mm0 \n\t"/*((a-b)/4-b)/4 */\ + "paddw %%mm6, %%mm2 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "psraw $6, %%mm0 \n\t"\ + "packuswb %%mm0, %%mm0 \n\t"\ OP(%%mm0, (%1),%%mm7, d)\ - "add $24, %0 \n\t"\ - "add %3, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add $24, %0 \n\t"\ + "add %3, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(tmp), "+c"(dst), "+m"(h)\ : "S"((long)dstStride), "m"(ff_pw_32)\ : "memory"\ @@ -537,54 +537,54 @@ static void OPNAME ## h264_qpel8_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\ int h=8;\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movq %5, %%mm6 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq 1(%0), %%mm2 \n\t"\ - "movq %%mm0, %%mm1 \n\t"\ - "movq %%mm2, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "paddw %%mm3, %%mm1 \n\t"\ - "psllw $2, %%mm0 \n\t"\ - "psllw $2, %%mm1 \n\t"\ - "movq -1(%0), %%mm2 \n\t"\ - "movq 2(%0), %%mm4 \n\t"\ - "movq %%mm2, %%mm3 \n\t"\ - "movq %%mm4, %%mm5 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ - "punpckhbw %%mm7, %%mm5 \n\t"\ - "paddw %%mm4, %%mm2 \n\t"\ - "paddw %%mm3, %%mm5 \n\t"\ - "psubw %%mm2, %%mm0 \n\t"\ - "psubw %%mm5, %%mm1 \n\t"\ - "pmullw %%mm6, %%mm0 \n\t"\ - "pmullw %%mm6, %%mm1 \n\t"\ - "movd -2(%0), %%mm2 \n\t"\ - "movd 7(%0), %%mm5 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm5 \n\t"\ - "paddw %%mm3, %%mm2 \n\t"\ - "paddw %%mm5, %%mm4 \n\t"\ - "movq %6, %%mm5 \n\t"\ - "paddw %%mm5, %%mm2 \n\t"\ - "paddw %%mm5, %%mm4 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "paddw %%mm4, %%mm1 \n\t"\ - "psraw $5, %%mm0 \n\t"\ - "psraw $5, %%mm1 \n\t"\ - "packuswb %%mm1, %%mm0 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movq %5, %%mm6 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 1(%0), %%mm2 \n\t"\ + "movq %%mm0, %%mm1 \n\t"\ + "movq %%mm2, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "paddw %%mm3, %%mm1 \n\t"\ + "psllw $2, %%mm0 \n\t"\ + "psllw $2, %%mm1 \n\t"\ + "movq -1(%0), %%mm2 \n\t"\ + "movq 2(%0), %%mm4 \n\t"\ + "movq %%mm2, %%mm3 \n\t"\ + "movq %%mm4, %%mm5 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ + "punpckhbw %%mm7, %%mm5 \n\t"\ + "paddw %%mm4, %%mm2 \n\t"\ + "paddw %%mm3, %%mm5 \n\t"\ + "psubw %%mm2, %%mm0 \n\t"\ + "psubw %%mm5, %%mm1 \n\t"\ + "pmullw %%mm6, %%mm0 \n\t"\ + "pmullw %%mm6, %%mm1 \n\t"\ + "movd -2(%0), %%mm2 \n\t"\ + "movd 7(%0), %%mm5 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm5 \n\t"\ + "paddw %%mm3, %%mm2 \n\t"\ + "paddw %%mm5, %%mm4 \n\t"\ + "movq %6, %%mm5 \n\t"\ + "paddw %%mm5, %%mm2 \n\t"\ + "paddw %%mm5, %%mm4 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "paddw %%mm4, %%mm1 \n\t"\ + "psraw $5, %%mm0 \n\t"\ + "psraw $5, %%mm1 \n\t"\ + "packuswb %%mm1, %%mm0 \n\t"\ OP(%%mm0, (%1),%%mm5, q)\ - "add %3, %0 \n\t"\ - "add %4, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add %3, %0 \n\t"\ + "add %4, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(src), "+c"(dst), "+m"(h)\ : "d"((long)srcStride), "S"((long)dstStride), "m"(ff_pw_5), "m"(ff_pw_16)\ : "memory"\ @@ -597,22 +597,22 @@ \ while(h--){\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movd (%0), %%mm0 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm1 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm2 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm3 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm4 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movd (%0), %%mm0 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm1 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm2 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm3 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm4 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ QPEL_H264V(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\ QPEL_H264V(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\ QPEL_H264V(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\ @@ -636,22 +636,22 @@ src -= 2*srcStride+2;\ while(w--){\ asm volatile(\ - "pxor %%mm7, %%mm7 \n\t"\ - "movd (%0), %%mm0 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm1 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm2 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm3 \n\t"\ - "add %2, %0 \n\t"\ - "movd (%0), %%mm4 \n\t"\ - "add %2, %0 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ + "pxor %%mm7, %%mm7 \n\t"\ + "movd (%0), %%mm0 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm1 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm2 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm3 \n\t"\ + "add %2, %0 \n\t"\ + "movd (%0), %%mm4 \n\t"\ + "add %2, %0 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ QPEL_H264HV(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, 0*8*4)\ QPEL_H264HV(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, 1*8*4)\ QPEL_H264HV(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, 2*8*4)\ @@ -670,42 +670,42 @@ }\ tmp -= 4*4;\ asm volatile(\ - "movq %4, %%mm6 \n\t"\ - "1: \n\t"\ - "movq (%0), %%mm0 \n\t"\ - "movq 8(%0), %%mm3 \n\t"\ - "movq 2(%0), %%mm1 \n\t"\ - "movq 10(%0), %%mm4 \n\t"\ - "paddw %%mm4, %%mm0 \n\t"\ - "paddw %%mm3, %%mm1 \n\t"\ - "paddw 18(%0), %%mm3 \n\t"\ - "paddw 16(%0), %%mm4 \n\t"\ - "movq 4(%0), %%mm2 \n\t"\ - "movq 12(%0), %%mm5 \n\t"\ - "paddw 6(%0), %%mm2 \n\t"\ - "paddw 14(%0), %%mm5 \n\t"\ - "psubw %%mm1, %%mm0 \n\t"\ - "psubw %%mm4, %%mm3 \n\t"\ - "psraw $2, %%mm0 \n\t"\ - "psraw $2, %%mm3 \n\t"\ - "psubw %%mm1, %%mm0 \n\t"\ - "psubw %%mm4, %%mm3 \n\t"\ - "paddsw %%mm2, %%mm0 \n\t"\ - "paddsw %%mm5, %%mm3 \n\t"\ - "psraw $2, %%mm0 \n\t"\ - "psraw $2, %%mm3 \n\t"\ - "paddw %%mm6, %%mm2 \n\t"\ - "paddw %%mm6, %%mm5 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "paddw %%mm5, %%mm3 \n\t"\ - "psraw $6, %%mm0 \n\t"\ - "psraw $6, %%mm3 \n\t"\ - "packuswb %%mm3, %%mm0 \n\t"\ + "movq %4, %%mm6 \n\t"\ + "1: \n\t"\ + "movq (%0), %%mm0 \n\t"\ + "movq 8(%0), %%mm3 \n\t"\ + "movq 2(%0), %%mm1 \n\t"\ + "movq 10(%0), %%mm4 \n\t"\ + "paddw %%mm4, %%mm0 \n\t"\ + "paddw %%mm3, %%mm1 \n\t"\ + "paddw 18(%0), %%mm3 \n\t"\ + "paddw 16(%0), %%mm4 \n\t"\ + "movq 4(%0), %%mm2 \n\t"\ + "movq 12(%0), %%mm5 \n\t"\ + "paddw 6(%0), %%mm2 \n\t"\ + "paddw 14(%0), %%mm5 \n\t"\ + "psubw %%mm1, %%mm0 \n\t"\ + "psubw %%mm4, %%mm3 \n\t"\ + "psraw $2, %%mm0 \n\t"\ + "psraw $2, %%mm3 \n\t"\ + "psubw %%mm1, %%mm0 \n\t"\ + "psubw %%mm4, %%mm3 \n\t"\ + "paddsw %%mm2, %%mm0 \n\t"\ + "paddsw %%mm5, %%mm3 \n\t"\ + "psraw $2, %%mm0 \n\t"\ + "psraw $2, %%mm3 \n\t"\ + "paddw %%mm6, %%mm2 \n\t"\ + "paddw %%mm6, %%mm5 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "paddw %%mm5, %%mm3 \n\t"\ + "psraw $6, %%mm0 \n\t"\ + "psraw $6, %%mm3 \n\t"\ + "packuswb %%mm3, %%mm0 \n\t"\ OP(%%mm0, (%1),%%mm7, q)\ - "add $32, %0 \n\t"\ - "add %3, %1 \n\t"\ - "decl %2 \n\t"\ - " jnz 1b \n\t"\ + "add $32, %0 \n\t"\ + "add %3, %1 \n\t"\ + "decl %2 \n\t"\ + " jnz 1b \n\t"\ : "+a"(tmp), "+c"(dst), "+m"(h)\ : "S"((long)dstStride), "m"(ff_pw_32)\ : "memory"\ @@ -862,15 +862,15 @@ }\ -#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t" +#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t" #define AVG_3DNOW_OP(a,b,temp, size) \ -"mov" #size " " #b ", " #temp " \n\t"\ -"pavgusb " #temp ", " #a " \n\t"\ -"mov" #size " " #a ", " #b " \n\t" +"mov" #size " " #b ", " #temp " \n\t"\ +"pavgusb " #temp ", " #a " \n\t"\ +"mov" #size " " #a ", " #b " \n\t" #define AVG_MMX2_OP(a,b,temp, size) \ -"mov" #size " " #b ", " #temp " \n\t"\ -"pavgb " #temp ", " #a " \n\t"\ -"mov" #size " " #a ", " #b " \n\t" +"mov" #size " " #b ", " #temp " \n\t"\ +"pavgb " #temp ", " #a " \n\t"\ +"mov" #size " " #a ", " #b " \n\t" QPEL_H264(put_, PUT_OP, 3dnow) QPEL_H264(avg_, AVG_3DNOW_OP, 3dnow)
--- a/i386/idct_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/idct_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -38,7 +38,7 @@ #if 0 /* C row IDCT - its just here to document the MMXEXT and MMX versions */ static inline void idct_row (int16_t * row, int offset, - int16_t * table, int32_t * rounder) + int16_t * table, int32_t * rounder) { int C1, C2, C3, C4, C5, C6, C7; int a0, a1, a2, a3, b0, b1, b2, b3; @@ -77,241 +77,241 @@ /* MMXEXT row IDCT */ -#define mmxext_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, -c4, -c2, \ - c4, c6, c4, c6, \ - c1, c3, -c1, -c5, \ - c5, c7, c3, -c7, \ - c4, -c6, c4, -c6, \ - -c4, c2, c4, -c2, \ - c5, -c1, c3, -c1, \ - c7, c3, c7, -c5 } +#define mmxext_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, -c4, -c2, \ + c4, c6, c4, c6, \ + c1, c3, -c1, -c5, \ + c5, c7, c3, -c7, \ + c4, -c6, c4, -c6, \ + -c4, c2, c4, -c2, \ + c5, -c1, c3, -c1, \ + c7, c3, c7, -c5 } static inline void mmxext_row_head (int16_t * row, int offset, const int16_t * table) { - movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 + movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 - movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 - movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 + movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 + movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 - movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4 - movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 + movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4 + movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 - movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4 - pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2 + movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4 + pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2 - pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4 + pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4 } static inline void mmxext_row (const int16_t * table, const int32_t * rounder) { - movq_m2r (*(table+8), mm1); // mm1 = -C5 -C1 C3 C1 - pmaddwd_r2r (mm2, mm4); // mm4 = C4*x0+C6*x2 C4*x4+C6*x6 + movq_m2r (*(table+8), mm1); // mm1 = -C5 -C1 C3 C1 + pmaddwd_r2r (mm2, mm4); // mm4 = C4*x0+C6*x2 C4*x4+C6*x6 - pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x4-C6*x6 C4*x0-C6*x2 - pshufw_r2r (mm6, mm6, 0x4e); // mm6 = x3 x1 x7 x5 + pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x4-C6*x6 C4*x0-C6*x2 + pshufw_r2r (mm6, mm6, 0x4e); // mm6 = x3 x1 x7 x5 - movq_m2r (*(table+12), mm7); // mm7 = -C7 C3 C7 C5 - pmaddwd_r2r (mm5, mm1); // mm1 = -C1*x5-C5*x7 C1*x1+C3*x3 + movq_m2r (*(table+12), mm7); // mm7 = -C7 C3 C7 C5 + pmaddwd_r2r (mm5, mm1); // mm1 = -C1*x5-C5*x7 C1*x1+C3*x3 - paddd_m2r (*rounder, mm3); // mm3 += rounder - pmaddwd_r2r (mm6, mm7); // mm7 = C3*x1-C7*x3 C5*x5+C7*x7 + paddd_m2r (*rounder, mm3); // mm3 += rounder + pmaddwd_r2r (mm6, mm7); // mm7 = C3*x1-C7*x3 C5*x5+C7*x7 - pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x0-C2*x2 -C4*x4+C2*x6 - paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder + pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x0-C2*x2 -C4*x4+C2*x6 + paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder - pmaddwd_m2r (*(table+24), mm5); // mm5 = C3*x5-C1*x7 C5*x1-C1*x3 - movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder + pmaddwd_m2r (*(table+24), mm5); // mm5 = C3*x5-C1*x7 C5*x1-C1*x3 + movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder - pmaddwd_m2r (*(table+28), mm6); // mm6 = C7*x1-C5*x3 C7*x5+C3*x7 - paddd_r2r (mm7, mm1); // mm1 = b1 b0 + pmaddwd_m2r (*(table+28), mm6); // mm6 = C7*x1-C5*x3 C7*x5+C3*x7 + paddd_r2r (mm7, mm1); // mm1 = b1 b0 - paddd_m2r (*rounder, mm0); // mm0 += rounder - psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder + paddd_m2r (*rounder, mm0); // mm0 += rounder + psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder - psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7 - paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder + psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7 + paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder - paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder - psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0 + paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder + psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0 - paddd_r2r (mm6, mm5); // mm5 = b3 b2 - movq_r2r (mm0, mm4); // mm4 = a3 a2 + rounder + paddd_r2r (mm6, mm5); // mm5 = b3 b2 + movq_r2r (mm0, mm4); // mm4 = a3 a2 + rounder - paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder - psubd_r2r (mm5, mm4); // mm4 = a3-b3 a2-b2 + rounder + paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder + psubd_r2r (mm5, mm4); // mm4 = a3-b3 a2-b2 + rounder } static inline void mmxext_row_tail (int16_t * row, int store) { - psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 + psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 - psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5 + psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5 - packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 + packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 - packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5 + packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5 - movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 - pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4 + movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 + pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4 /* slot */ - movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4 + movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4 } static inline void mmxext_row_mid (int16_t * row, int store, - int offset, const int16_t * table) + int offset, const int16_t * table) { - movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 - psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 + movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 + psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 - movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 - psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5 + movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 + psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5 - packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 - movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 + packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 + movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 - packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5 - movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 + packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5 + movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 - movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 - pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4 + movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 + pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4 - movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4 - movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4 + movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4 + movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4 - pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2 + pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2 - movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4 - pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4 + movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4 + pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4 } /* MMX row IDCT */ -#define mmx_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, c4, c6, \ - c4, c6, -c4, -c2, \ - c1, c3, c3, -c7, \ - c5, c7, -c1, -c5, \ - c4, -c6, c4, -c2, \ - -c4, c2, c4, -c6, \ - c5, -c1, c7, -c5, \ - c7, c3, c3, -c1 } +#define mmx_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, c4, c6, \ + c4, c6, -c4, -c2, \ + c1, c3, c3, -c7, \ + c5, c7, -c1, -c5, \ + c4, -c6, c4, -c2, \ + -c4, c2, c4, -c6, \ + c5, -c1, c7, -c5, \ + c7, c3, c3, -c1 } static inline void mmx_row_head (int16_t * row, int offset, const int16_t * table) { - movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 + movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 - movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 - movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 + movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 + movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 - movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4 - movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 + movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4 + movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 - punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0 + punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0 - movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4 - pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2 + movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4 + pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2 - movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1 - punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4 + movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1 + punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4 } static inline void mmx_row (const int16_t * table, const int32_t * rounder) { - pmaddwd_r2r (mm2, mm4); // mm4 = -C4*x4-C2*x6 C4*x4+C6*x6 - punpckldq_r2r (mm5, mm5); // mm5 = x3 x1 x3 x1 + pmaddwd_r2r (mm2, mm4); // mm4 = -C4*x4-C2*x6 C4*x4+C6*x6 + punpckldq_r2r (mm5, mm5); // mm5 = x3 x1 x3 x1 - pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x0-C2*x2 C4*x0-C6*x2 - punpckhdq_r2r (mm6, mm6); // mm6 = x7 x5 x7 x5 + pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x0-C2*x2 C4*x0-C6*x2 + punpckhdq_r2r (mm6, mm6); // mm6 = x7 x5 x7 x5 - movq_m2r (*(table+12), mm7); // mm7 = -C5 -C1 C7 C5 - pmaddwd_r2r (mm5, mm1); // mm1 = C3*x1-C7*x3 C1*x1+C3*x3 + movq_m2r (*(table+12), mm7); // mm7 = -C5 -C1 C7 C5 + pmaddwd_r2r (mm5, mm1); // mm1 = C3*x1-C7*x3 C1*x1+C3*x3 - paddd_m2r (*rounder, mm3); // mm3 += rounder - pmaddwd_r2r (mm6, mm7); // mm7 = -C1*x5-C5*x7 C5*x5+C7*x7 + paddd_m2r (*rounder, mm3); // mm3 += rounder + pmaddwd_r2r (mm6, mm7); // mm7 = -C1*x5-C5*x7 C5*x5+C7*x7 - pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x4-C6*x6 -C4*x4+C2*x6 - paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder + pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x4-C6*x6 -C4*x4+C2*x6 + paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder - pmaddwd_m2r (*(table+24), mm5); // mm5 = C7*x1-C5*x3 C5*x1-C1*x3 - movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder + pmaddwd_m2r (*(table+24), mm5); // mm5 = C7*x1-C5*x3 C5*x1-C1*x3 + movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder - pmaddwd_m2r (*(table+28), mm6); // mm6 = C3*x5-C1*x7 C7*x5+C3*x7 - paddd_r2r (mm7, mm1); // mm1 = b1 b0 + pmaddwd_m2r (*(table+28), mm6); // mm6 = C3*x5-C1*x7 C7*x5+C3*x7 + paddd_r2r (mm7, mm1); // mm1 = b1 b0 - paddd_m2r (*rounder, mm0); // mm0 += rounder - psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder + paddd_m2r (*rounder, mm0); // mm0 += rounder + psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder - psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7 - paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder + psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7 + paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder - paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder - psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0 + paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder + psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0 - paddd_r2r (mm6, mm5); // mm5 = b3 b2 - movq_r2r (mm0, mm7); // mm7 = a3 a2 + rounder + paddd_r2r (mm6, mm5); // mm5 = b3 b2 + movq_r2r (mm0, mm7); // mm7 = a3 a2 + rounder - paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder - psubd_r2r (mm5, mm7); // mm7 = a3-b3 a2-b2 + rounder + paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder + psubd_r2r (mm5, mm7); // mm7 = a3-b3 a2-b2 + rounder } static inline void mmx_row_tail (int16_t * row, int store) { - psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 + psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 - psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5 + psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5 - packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 + packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 - packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5 + packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5 - movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 - movq_r2r (mm7, mm4); // mm4 = y6 y7 y4 y5 + movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 + movq_r2r (mm7, mm4); // mm4 = y6 y7 y4 y5 - pslld_i2r (16, mm7); // mm7 = y7 0 y5 0 + pslld_i2r (16, mm7); // mm7 = y7 0 y5 0 - psrld_i2r (16, mm4); // mm4 = 0 y6 0 y4 + psrld_i2r (16, mm4); // mm4 = 0 y6 0 y4 - por_r2r (mm4, mm7); // mm7 = y7 y6 y5 y4 + por_r2r (mm4, mm7); // mm7 = y7 y6 y5 y4 /* slot */ - movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4 + movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4 } static inline void mmx_row_mid (int16_t * row, int store, - int offset, const int16_t * table) + int offset, const int16_t * table) { - movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 - psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 + movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0 + psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2 - movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 - psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5 + movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1 + psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5 - packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 - movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 + packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0 + movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1 - packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5 - movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 + packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5 + movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0 - movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 - movq_r2r (mm7, mm1); // mm1 = y6 y7 y4 y5 + movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0 + movq_r2r (mm7, mm1); // mm1 = y6 y7 y4 y5 - punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0 - psrld_i2r (16, mm7); // mm7 = 0 y6 0 y4 + punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0 + psrld_i2r (16, mm7); // mm7 = 0 y6 0 y4 - movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4 - pslld_i2r (16, mm1); // mm1 = y7 0 y5 0 + movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4 + pslld_i2r (16, mm1); // mm1 = y7 0 y5 0 - movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4 - por_r2r (mm1, mm7); // mm7 = y7 y6 y5 y4 + movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4 + por_r2r (mm1, mm7); // mm7 = y7 y6 y5 y4 - movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1 - punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4 + movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1 + punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4 - movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4 - pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2 + movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4 + pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2 } @@ -403,132 +403,132 @@ /* column code adapted from peter gubanov */ /* http://www.elecard.com/peter/idct.shtml */ - movq_m2r (*_T1, mm0); // mm0 = T1 + movq_m2r (*_T1, mm0); // mm0 = T1 - movq_m2r (*(col+offset+1*8), mm1); // mm1 = x1 - movq_r2r (mm0, mm2); // mm2 = T1 + movq_m2r (*(col+offset+1*8), mm1); // mm1 = x1 + movq_r2r (mm0, mm2); // mm2 = T1 - movq_m2r (*(col+offset+7*8), mm4); // mm4 = x7 - pmulhw_r2r (mm1, mm0); // mm0 = T1*x1 + movq_m2r (*(col+offset+7*8), mm4); // mm4 = x7 + pmulhw_r2r (mm1, mm0); // mm0 = T1*x1 - movq_m2r (*_T3, mm5); // mm5 = T3 - pmulhw_r2r (mm4, mm2); // mm2 = T1*x7 + movq_m2r (*_T3, mm5); // mm5 = T3 + pmulhw_r2r (mm4, mm2); // mm2 = T1*x7 - movq_m2r (*(col+offset+5*8), mm6); // mm6 = x5 - movq_r2r (mm5, mm7); // mm7 = T3-1 + movq_m2r (*(col+offset+5*8), mm6); // mm6 = x5 + movq_r2r (mm5, mm7); // mm7 = T3-1 - movq_m2r (*(col+offset+3*8), mm3); // mm3 = x3 - psubsw_r2r (mm4, mm0); // mm0 = v17 + movq_m2r (*(col+offset+3*8), mm3); // mm3 = x3 + psubsw_r2r (mm4, mm0); // mm0 = v17 - movq_m2r (*_T2, mm4); // mm4 = T2 - pmulhw_r2r (mm3, mm5); // mm5 = (T3-1)*x3 + movq_m2r (*_T2, mm4); // mm4 = T2 + pmulhw_r2r (mm3, mm5); // mm5 = (T3-1)*x3 - paddsw_r2r (mm2, mm1); // mm1 = u17 - pmulhw_r2r (mm6, mm7); // mm7 = (T3-1)*x5 + paddsw_r2r (mm2, mm1); // mm1 = u17 + pmulhw_r2r (mm6, mm7); // mm7 = (T3-1)*x5 /* slot */ - movq_r2r (mm4, mm2); // mm2 = T2 - paddsw_r2r (mm3, mm5); // mm5 = T3*x3 + movq_r2r (mm4, mm2); // mm2 = T2 + paddsw_r2r (mm3, mm5); // mm5 = T3*x3 pmulhw_m2r (*(col+offset+2*8), mm4);// mm4 = T2*x2 - paddsw_r2r (mm6, mm7); // mm7 = T3*x5 + paddsw_r2r (mm6, mm7); // mm7 = T3*x5 - psubsw_r2r (mm6, mm5); // mm5 = v35 - paddsw_r2r (mm3, mm7); // mm7 = u35 + psubsw_r2r (mm6, mm5); // mm5 = v35 + paddsw_r2r (mm3, mm7); // mm7 = u35 - movq_m2r (*(col+offset+6*8), mm3); // mm3 = x6 - movq_r2r (mm0, mm6); // mm6 = v17 + movq_m2r (*(col+offset+6*8), mm3); // mm3 = x6 + movq_r2r (mm0, mm6); // mm6 = v17 - pmulhw_r2r (mm3, mm2); // mm2 = T2*x6 - psubsw_r2r (mm5, mm0); // mm0 = b3 + pmulhw_r2r (mm3, mm2); // mm2 = T2*x6 + psubsw_r2r (mm5, mm0); // mm0 = b3 - psubsw_r2r (mm3, mm4); // mm4 = v26 - paddsw_r2r (mm6, mm5); // mm5 = v12 + psubsw_r2r (mm3, mm4); // mm4 = v26 + paddsw_r2r (mm6, mm5); // mm5 = v12 - movq_r2m (mm0, *(col+offset+3*8)); // save b3 in scratch0 - movq_r2r (mm1, mm6); // mm6 = u17 + movq_r2m (mm0, *(col+offset+3*8)); // save b3 in scratch0 + movq_r2r (mm1, mm6); // mm6 = u17 paddsw_m2r (*(col+offset+2*8), mm2);// mm2 = u26 - paddsw_r2r (mm7, mm6); // mm6 = b0 + paddsw_r2r (mm7, mm6); // mm6 = b0 - psubsw_r2r (mm7, mm1); // mm1 = u12 - movq_r2r (mm1, mm7); // mm7 = u12 + psubsw_r2r (mm7, mm1); // mm1 = u12 + movq_r2r (mm1, mm7); // mm7 = u12 - movq_m2r (*(col+offset+0*8), mm3); // mm3 = x0 - paddsw_r2r (mm5, mm1); // mm1 = u12+v12 + movq_m2r (*(col+offset+0*8), mm3); // mm3 = x0 + paddsw_r2r (mm5, mm1); // mm1 = u12+v12 - movq_m2r (*_C4, mm0); // mm0 = C4/2 - psubsw_r2r (mm5, mm7); // mm7 = u12-v12 + movq_m2r (*_C4, mm0); // mm0 = C4/2 + psubsw_r2r (mm5, mm7); // mm7 = u12-v12 - movq_r2m (mm6, *(col+offset+5*8)); // save b0 in scratch1 - pmulhw_r2r (mm0, mm1); // mm1 = b1/2 + movq_r2m (mm6, *(col+offset+5*8)); // save b0 in scratch1 + pmulhw_r2r (mm0, mm1); // mm1 = b1/2 - movq_r2r (mm4, mm6); // mm6 = v26 - pmulhw_r2r (mm0, mm7); // mm7 = b2/2 + movq_r2r (mm4, mm6); // mm6 = v26 + pmulhw_r2r (mm0, mm7); // mm7 = b2/2 - movq_m2r (*(col+offset+4*8), mm5); // mm5 = x4 - movq_r2r (mm3, mm0); // mm0 = x0 + movq_m2r (*(col+offset+4*8), mm5); // mm5 = x4 + movq_r2r (mm3, mm0); // mm0 = x0 - psubsw_r2r (mm5, mm3); // mm3 = v04 - paddsw_r2r (mm5, mm0); // mm0 = u04 + psubsw_r2r (mm5, mm3); // mm3 = v04 + paddsw_r2r (mm5, mm0); // mm0 = u04 - paddsw_r2r (mm3, mm4); // mm4 = a1 - movq_r2r (mm0, mm5); // mm5 = u04 + paddsw_r2r (mm3, mm4); // mm4 = a1 + movq_r2r (mm0, mm5); // mm5 = u04 - psubsw_r2r (mm6, mm3); // mm3 = a2 - paddsw_r2r (mm2, mm5); // mm5 = a0 + psubsw_r2r (mm6, mm3); // mm3 = a2 + paddsw_r2r (mm2, mm5); // mm5 = a0 - paddsw_r2r (mm1, mm1); // mm1 = b1 - psubsw_r2r (mm2, mm0); // mm0 = a3 + paddsw_r2r (mm1, mm1); // mm1 = b1 + psubsw_r2r (mm2, mm0); // mm0 = a3 - paddsw_r2r (mm7, mm7); // mm7 = b2 - movq_r2r (mm3, mm2); // mm2 = a2 + paddsw_r2r (mm7, mm7); // mm7 = b2 + movq_r2r (mm3, mm2); // mm2 = a2 - movq_r2r (mm4, mm6); // mm6 = a1 - paddsw_r2r (mm7, mm3); // mm3 = a2+b2 + movq_r2r (mm4, mm6); // mm6 = a1 + paddsw_r2r (mm7, mm3); // mm3 = a2+b2 - psraw_i2r (COL_SHIFT, mm3); // mm3 = y2 - paddsw_r2r (mm1, mm4); // mm4 = a1+b1 + psraw_i2r (COL_SHIFT, mm3); // mm3 = y2 + paddsw_r2r (mm1, mm4); // mm4 = a1+b1 - psraw_i2r (COL_SHIFT, mm4); // mm4 = y1 - psubsw_r2r (mm1, mm6); // mm6 = a1-b1 + psraw_i2r (COL_SHIFT, mm4); // mm4 = y1 + psubsw_r2r (mm1, mm6); // mm6 = a1-b1 - movq_m2r (*(col+offset+5*8), mm1); // mm1 = b0 - psubsw_r2r (mm7, mm2); // mm2 = a2-b2 + movq_m2r (*(col+offset+5*8), mm1); // mm1 = b0 + psubsw_r2r (mm7, mm2); // mm2 = a2-b2 - psraw_i2r (COL_SHIFT, mm6); // mm6 = y6 - movq_r2r (mm5, mm7); // mm7 = a0 + psraw_i2r (COL_SHIFT, mm6); // mm6 = y6 + movq_r2r (mm5, mm7); // mm7 = a0 - movq_r2m (mm4, *(col+offset+1*8)); // save y1 - psraw_i2r (COL_SHIFT, mm2); // mm2 = y5 + movq_r2m (mm4, *(col+offset+1*8)); // save y1 + psraw_i2r (COL_SHIFT, mm2); // mm2 = y5 - movq_r2m (mm3, *(col+offset+2*8)); // save y2 - paddsw_r2r (mm1, mm5); // mm5 = a0+b0 + movq_r2m (mm3, *(col+offset+2*8)); // save y2 + paddsw_r2r (mm1, mm5); // mm5 = a0+b0 - movq_m2r (*(col+offset+3*8), mm4); // mm4 = b3 - psubsw_r2r (mm1, mm7); // mm7 = a0-b0 + movq_m2r (*(col+offset+3*8), mm4); // mm4 = b3 + psubsw_r2r (mm1, mm7); // mm7 = a0-b0 - psraw_i2r (COL_SHIFT, mm5); // mm5 = y0 - movq_r2r (mm0, mm3); // mm3 = a3 + psraw_i2r (COL_SHIFT, mm5); // mm5 = y0 + movq_r2r (mm0, mm3); // mm3 = a3 - movq_r2m (mm2, *(col+offset+5*8)); // save y5 - psubsw_r2r (mm4, mm3); // mm3 = a3-b3 + movq_r2m (mm2, *(col+offset+5*8)); // save y5 + psubsw_r2r (mm4, mm3); // mm3 = a3-b3 - psraw_i2r (COL_SHIFT, mm7); // mm7 = y7 - paddsw_r2r (mm0, mm4); // mm4 = a3+b3 + psraw_i2r (COL_SHIFT, mm7); // mm7 = y7 + paddsw_r2r (mm0, mm4); // mm4 = a3+b3 - movq_r2m (mm5, *(col+offset+0*8)); // save y0 - psraw_i2r (COL_SHIFT, mm3); // mm3 = y4 + movq_r2m (mm5, *(col+offset+0*8)); // save y0 + psraw_i2r (COL_SHIFT, mm3); // mm3 = y4 - movq_r2m (mm6, *(col+offset+6*8)); // save y6 - psraw_i2r (COL_SHIFT, mm4); // mm4 = y3 + movq_r2m (mm6, *(col+offset+6*8)); // save y6 + psraw_i2r (COL_SHIFT, mm4); // mm4 = y3 - movq_r2m (mm7, *(col+offset+7*8)); // save y7 + movq_r2m (mm7, *(col+offset+7*8)); // save y7 - movq_r2m (mm3, *(col+offset+4*8)); // save y4 + movq_r2m (mm3, *(col+offset+4*8)); // save y4 - movq_r2m (mm4, *(col+offset+3*8)); // save y3 + movq_r2m (mm4, *(col+offset+3*8)); // save y3 #undef T1 #undef T2 @@ -540,61 +540,61 @@ rounder ((1 << (COL_SHIFT - 1)) - 0.5); static const int32_t rounder4[] ATTR_ALIGN(8) = rounder (0); static const int32_t rounder1[] ATTR_ALIGN(8) = - rounder (1.25683487303); /* C1*(C1/C4+C1+C7)/2 */ + rounder (1.25683487303); /* C1*(C1/C4+C1+C7)/2 */ static const int32_t rounder7[] ATTR_ALIGN(8) = - rounder (-0.25); /* C1*(C7/C4+C7-C1)/2 */ + rounder (-0.25); /* C1*(C7/C4+C7-C1)/2 */ static const int32_t rounder2[] ATTR_ALIGN(8) = - rounder (0.60355339059); /* C2 * (C6+C2)/2 */ + rounder (0.60355339059); /* C2 * (C6+C2)/2 */ static const int32_t rounder6[] ATTR_ALIGN(8) = - rounder (-0.25); /* C2 * (C6-C2)/2 */ + rounder (-0.25); /* C2 * (C6-C2)/2 */ static const int32_t rounder3[] ATTR_ALIGN(8) = - rounder (0.087788325588); /* C3*(-C3/C4+C3+C5)/2 */ + rounder (0.087788325588); /* C3*(-C3/C4+C3+C5)/2 */ static const int32_t rounder5[] ATTR_ALIGN(8) = - rounder (-0.441341716183); /* C3*(-C5/C4+C5-C3)/2 */ + rounder (-0.441341716183); /* C3*(-C5/C4+C5-C3)/2 */ #undef COL_SHIFT #undef ROW_SHIFT -#define declare_idct(idct,table,idct_row_head,idct_row,idct_row_tail,idct_row_mid) \ -void idct (int16_t * block) \ -{ \ - static const int16_t table04[] ATTR_ALIGN(16) = \ - table (22725, 21407, 19266, 16384, 12873, 8867, 4520); \ - static const int16_t table17[] ATTR_ALIGN(16) = \ - table (31521, 29692, 26722, 22725, 17855, 12299, 6270); \ - static const int16_t table26[] ATTR_ALIGN(16) = \ - table (29692, 27969, 25172, 21407, 16819, 11585, 5906); \ - static const int16_t table35[] ATTR_ALIGN(16) = \ - table (26722, 25172, 22654, 19266, 15137, 10426, 5315); \ - \ - idct_row_head (block, 0*8, table04); \ - idct_row (table04, rounder0); \ - idct_row_mid (block, 0*8, 4*8, table04); \ - idct_row (table04, rounder4); \ - idct_row_mid (block, 4*8, 1*8, table17); \ - idct_row (table17, rounder1); \ - idct_row_mid (block, 1*8, 7*8, table17); \ - idct_row (table17, rounder7); \ - idct_row_mid (block, 7*8, 2*8, table26); \ - idct_row (table26, rounder2); \ - idct_row_mid (block, 2*8, 6*8, table26); \ - idct_row (table26, rounder6); \ - idct_row_mid (block, 6*8, 3*8, table35); \ - idct_row (table35, rounder3); \ - idct_row_mid (block, 3*8, 5*8, table35); \ - idct_row (table35, rounder5); \ - idct_row_tail (block, 5*8); \ - \ - idct_col (block, 0); \ - idct_col (block, 4); \ +#define declare_idct(idct,table,idct_row_head,idct_row,idct_row_tail,idct_row_mid) \ +void idct (int16_t * block) \ +{ \ + static const int16_t table04[] ATTR_ALIGN(16) = \ + table (22725, 21407, 19266, 16384, 12873, 8867, 4520); \ + static const int16_t table17[] ATTR_ALIGN(16) = \ + table (31521, 29692, 26722, 22725, 17855, 12299, 6270); \ + static const int16_t table26[] ATTR_ALIGN(16) = \ + table (29692, 27969, 25172, 21407, 16819, 11585, 5906); \ + static const int16_t table35[] ATTR_ALIGN(16) = \ + table (26722, 25172, 22654, 19266, 15137, 10426, 5315); \ + \ + idct_row_head (block, 0*8, table04); \ + idct_row (table04, rounder0); \ + idct_row_mid (block, 0*8, 4*8, table04); \ + idct_row (table04, rounder4); \ + idct_row_mid (block, 4*8, 1*8, table17); \ + idct_row (table17, rounder1); \ + idct_row_mid (block, 1*8, 7*8, table17); \ + idct_row (table17, rounder7); \ + idct_row_mid (block, 7*8, 2*8, table26); \ + idct_row (table26, rounder2); \ + idct_row_mid (block, 2*8, 6*8, table26); \ + idct_row (table26, rounder6); \ + idct_row_mid (block, 6*8, 3*8, table35); \ + idct_row (table35, rounder3); \ + idct_row_mid (block, 3*8, 5*8, table35); \ + idct_row (table35, rounder5); \ + idct_row_tail (block, 5*8); \ + \ + idct_col (block, 0); \ + idct_col (block, 4); \ } void ff_mmx_idct(DCTELEM *block); void ff_mmxext_idct(DCTELEM *block); declare_idct (ff_mmxext_idct, mmxext_table, - mmxext_row_head, mmxext_row, mmxext_row_tail, mmxext_row_mid) + mmxext_row_head, mmxext_row, mmxext_row_tail, mmxext_row_mid) declare_idct (ff_mmx_idct, mmx_table, - mmx_row_head, mmx_row, mmx_row_tail, mmx_row_mid) + mmx_row_head, mmx_row, mmx_row_tail, mmx_row_mid)
--- a/i386/mmx.h Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/mmx.h Thu Dec 22 01:10:11 2005 +0000 @@ -27,257 +27,257 @@ * values by ULL, lest they be truncated by the compiler) */ -typedef union { - long long q; /* Quadword (64-bit) value */ - unsigned long long uq; /* Unsigned Quadword */ - int d[2]; /* 2 Doubleword (32-bit) values */ - unsigned int ud[2]; /* 2 Unsigned Doubleword */ - short w[4]; /* 4 Word (16-bit) values */ - unsigned short uw[4]; /* 4 Unsigned Word */ - char b[8]; /* 8 Byte (8-bit) values */ - unsigned char ub[8]; /* 8 Unsigned Byte */ - float s[2]; /* Single-precision (32-bit) value */ -} mmx_t; /* On an 8-byte (64-bit) boundary */ +typedef union { + long long q; /* Quadword (64-bit) value */ + unsigned long long uq; /* Unsigned Quadword */ + int d[2]; /* 2 Doubleword (32-bit) values */ + unsigned int ud[2]; /* 2 Unsigned Doubleword */ + short w[4]; /* 4 Word (16-bit) values */ + unsigned short uw[4]; /* 4 Unsigned Word */ + char b[8]; /* 8 Byte (8-bit) values */ + unsigned char ub[8]; /* 8 Unsigned Byte */ + float s[2]; /* Single-precision (32-bit) value */ +} mmx_t; /* On an 8-byte (64-bit) boundary */ -#define mmx_i2r(op,imm,reg) \ - __asm__ __volatile__ (#op " %0, %%" #reg \ - : /* nothing */ \ - : "i" (imm) ) +#define mmx_i2r(op,imm,reg) \ + __asm__ __volatile__ (#op " %0, %%" #reg \ + : /* nothing */ \ + : "i" (imm) ) -#define mmx_m2r(op,mem,reg) \ - __asm__ __volatile__ (#op " %0, %%" #reg \ - : /* nothing */ \ - : "m" (mem)) +#define mmx_m2r(op,mem,reg) \ + __asm__ __volatile__ (#op " %0, %%" #reg \ + : /* nothing */ \ + : "m" (mem)) -#define mmx_r2m(op,reg,mem) \ - __asm__ __volatile__ (#op " %%" #reg ", %0" \ - : "=m" (mem) \ - : /* nothing */ ) +#define mmx_r2m(op,reg,mem) \ + __asm__ __volatile__ (#op " %%" #reg ", %0" \ + : "=m" (mem) \ + : /* nothing */ ) -#define mmx_r2r(op,regs,regd) \ - __asm__ __volatile__ (#op " %" #regs ", %" #regd) +#define mmx_r2r(op,regs,regd) \ + __asm__ __volatile__ (#op " %" #regs ", %" #regd) -#define emms() __asm__ __volatile__ ("emms") +#define emms() __asm__ __volatile__ ("emms") -#define movd_m2r(var,reg) mmx_m2r (movd, var, reg) -#define movd_r2m(reg,var) mmx_r2m (movd, reg, var) -#define movd_r2r(regs,regd) mmx_r2r (movd, regs, regd) +#define movd_m2r(var,reg) mmx_m2r (movd, var, reg) +#define movd_r2m(reg,var) mmx_r2m (movd, reg, var) +#define movd_r2r(regs,regd) mmx_r2r (movd, regs, regd) -#define movq_m2r(var,reg) mmx_m2r (movq, var, reg) -#define movq_r2m(reg,var) mmx_r2m (movq, reg, var) -#define movq_r2r(regs,regd) mmx_r2r (movq, regs, regd) +#define movq_m2r(var,reg) mmx_m2r (movq, var, reg) +#define movq_r2m(reg,var) mmx_r2m (movq, reg, var) +#define movq_r2r(regs,regd) mmx_r2r (movq, regs, regd) -#define packssdw_m2r(var,reg) mmx_m2r (packssdw, var, reg) -#define packssdw_r2r(regs,regd) mmx_r2r (packssdw, regs, regd) -#define packsswb_m2r(var,reg) mmx_m2r (packsswb, var, reg) -#define packsswb_r2r(regs,regd) mmx_r2r (packsswb, regs, regd) +#define packssdw_m2r(var,reg) mmx_m2r (packssdw, var, reg) +#define packssdw_r2r(regs,regd) mmx_r2r (packssdw, regs, regd) +#define packsswb_m2r(var,reg) mmx_m2r (packsswb, var, reg) +#define packsswb_r2r(regs,regd) mmx_r2r (packsswb, regs, regd) -#define packuswb_m2r(var,reg) mmx_m2r (packuswb, var, reg) -#define packuswb_r2r(regs,regd) mmx_r2r (packuswb, regs, regd) +#define packuswb_m2r(var,reg) mmx_m2r (packuswb, var, reg) +#define packuswb_r2r(regs,regd) mmx_r2r (packuswb, regs, regd) -#define paddb_m2r(var,reg) mmx_m2r (paddb, var, reg) -#define paddb_r2r(regs,regd) mmx_r2r (paddb, regs, regd) -#define paddd_m2r(var,reg) mmx_m2r (paddd, var, reg) -#define paddd_r2r(regs,regd) mmx_r2r (paddd, regs, regd) -#define paddw_m2r(var,reg) mmx_m2r (paddw, var, reg) -#define paddw_r2r(regs,regd) mmx_r2r (paddw, regs, regd) +#define paddb_m2r(var,reg) mmx_m2r (paddb, var, reg) +#define paddb_r2r(regs,regd) mmx_r2r (paddb, regs, regd) +#define paddd_m2r(var,reg) mmx_m2r (paddd, var, reg) +#define paddd_r2r(regs,regd) mmx_r2r (paddd, regs, regd) +#define paddw_m2r(var,reg) mmx_m2r (paddw, var, reg) +#define paddw_r2r(regs,regd) mmx_r2r (paddw, regs, regd) -#define paddsb_m2r(var,reg) mmx_m2r (paddsb, var, reg) -#define paddsb_r2r(regs,regd) mmx_r2r (paddsb, regs, regd) -#define paddsw_m2r(var,reg) mmx_m2r (paddsw, var, reg) -#define paddsw_r2r(regs,regd) mmx_r2r (paddsw, regs, regd) +#define paddsb_m2r(var,reg) mmx_m2r (paddsb, var, reg) +#define paddsb_r2r(regs,regd) mmx_r2r (paddsb, regs, regd) +#define paddsw_m2r(var,reg) mmx_m2r (paddsw, var, reg) +#define paddsw_r2r(regs,regd) mmx_r2r (paddsw, regs, regd) -#define paddusb_m2r(var,reg) mmx_m2r (paddusb, var, reg) -#define paddusb_r2r(regs,regd) mmx_r2r (paddusb, regs, regd) -#define paddusw_m2r(var,reg) mmx_m2r (paddusw, var, reg) -#define paddusw_r2r(regs,regd) mmx_r2r (paddusw, regs, regd) +#define paddusb_m2r(var,reg) mmx_m2r (paddusb, var, reg) +#define paddusb_r2r(regs,regd) mmx_r2r (paddusb, regs, regd) +#define paddusw_m2r(var,reg) mmx_m2r (paddusw, var, reg) +#define paddusw_r2r(regs,regd) mmx_r2r (paddusw, regs, regd) -#define pand_m2r(var,reg) mmx_m2r (pand, var, reg) -#define pand_r2r(regs,regd) mmx_r2r (pand, regs, regd) +#define pand_m2r(var,reg) mmx_m2r (pand, var, reg) +#define pand_r2r(regs,regd) mmx_r2r (pand, regs, regd) -#define pandn_m2r(var,reg) mmx_m2r (pandn, var, reg) -#define pandn_r2r(regs,regd) mmx_r2r (pandn, regs, regd) +#define pandn_m2r(var,reg) mmx_m2r (pandn, var, reg) +#define pandn_r2r(regs,regd) mmx_r2r (pandn, regs, regd) -#define pcmpeqb_m2r(var,reg) mmx_m2r (pcmpeqb, var, reg) -#define pcmpeqb_r2r(regs,regd) mmx_r2r (pcmpeqb, regs, regd) -#define pcmpeqd_m2r(var,reg) mmx_m2r (pcmpeqd, var, reg) -#define pcmpeqd_r2r(regs,regd) mmx_r2r (pcmpeqd, regs, regd) -#define pcmpeqw_m2r(var,reg) mmx_m2r (pcmpeqw, var, reg) -#define pcmpeqw_r2r(regs,regd) mmx_r2r (pcmpeqw, regs, regd) +#define pcmpeqb_m2r(var,reg) mmx_m2r (pcmpeqb, var, reg) +#define pcmpeqb_r2r(regs,regd) mmx_r2r (pcmpeqb, regs, regd) +#define pcmpeqd_m2r(var,reg) mmx_m2r (pcmpeqd, var, reg) +#define pcmpeqd_r2r(regs,regd) mmx_r2r (pcmpeqd, regs, regd) +#define pcmpeqw_m2r(var,reg) mmx_m2r (pcmpeqw, var, reg) +#define pcmpeqw_r2r(regs,regd) mmx_r2r (pcmpeqw, regs, regd) -#define pcmpgtb_m2r(var,reg) mmx_m2r (pcmpgtb, var, reg) -#define pcmpgtb_r2r(regs,regd) mmx_r2r (pcmpgtb, regs, regd) -#define pcmpgtd_m2r(var,reg) mmx_m2r (pcmpgtd, var, reg) -#define pcmpgtd_r2r(regs,regd) mmx_r2r (pcmpgtd, regs, regd) -#define pcmpgtw_m2r(var,reg) mmx_m2r (pcmpgtw, var, reg) -#define pcmpgtw_r2r(regs,regd) mmx_r2r (pcmpgtw, regs, regd) +#define pcmpgtb_m2r(var,reg) mmx_m2r (pcmpgtb, var, reg) +#define pcmpgtb_r2r(regs,regd) mmx_r2r (pcmpgtb, regs, regd) +#define pcmpgtd_m2r(var,reg) mmx_m2r (pcmpgtd, var, reg) +#define pcmpgtd_r2r(regs,regd) mmx_r2r (pcmpgtd, regs, regd) +#define pcmpgtw_m2r(var,reg) mmx_m2r (pcmpgtw, var, reg) +#define pcmpgtw_r2r(regs,regd) mmx_r2r (pcmpgtw, regs, regd) -#define pmaddwd_m2r(var,reg) mmx_m2r (pmaddwd, var, reg) -#define pmaddwd_r2r(regs,regd) mmx_r2r (pmaddwd, regs, regd) +#define pmaddwd_m2r(var,reg) mmx_m2r (pmaddwd, var, reg) +#define pmaddwd_r2r(regs,regd) mmx_r2r (pmaddwd, regs, regd) -#define pmulhw_m2r(var,reg) mmx_m2r (pmulhw, var, reg) -#define pmulhw_r2r(regs,regd) mmx_r2r (pmulhw, regs, regd) +#define pmulhw_m2r(var,reg) mmx_m2r (pmulhw, var, reg) +#define pmulhw_r2r(regs,regd) mmx_r2r (pmulhw, regs, regd) -#define pmullw_m2r(var,reg) mmx_m2r (pmullw, var, reg) -#define pmullw_r2r(regs,regd) mmx_r2r (pmullw, regs, regd) +#define pmullw_m2r(var,reg) mmx_m2r (pmullw, var, reg) +#define pmullw_r2r(regs,regd) mmx_r2r (pmullw, regs, regd) -#define por_m2r(var,reg) mmx_m2r (por, var, reg) -#define por_r2r(regs,regd) mmx_r2r (por, regs, regd) +#define por_m2r(var,reg) mmx_m2r (por, var, reg) +#define por_r2r(regs,regd) mmx_r2r (por, regs, regd) -#define pslld_i2r(imm,reg) mmx_i2r (pslld, imm, reg) -#define pslld_m2r(var,reg) mmx_m2r (pslld, var, reg) -#define pslld_r2r(regs,regd) mmx_r2r (pslld, regs, regd) -#define psllq_i2r(imm,reg) mmx_i2r (psllq, imm, reg) -#define psllq_m2r(var,reg) mmx_m2r (psllq, var, reg) -#define psllq_r2r(regs,regd) mmx_r2r (psllq, regs, regd) -#define psllw_i2r(imm,reg) mmx_i2r (psllw, imm, reg) -#define psllw_m2r(var,reg) mmx_m2r (psllw, var, reg) -#define psllw_r2r(regs,regd) mmx_r2r (psllw, regs, regd) +#define pslld_i2r(imm,reg) mmx_i2r (pslld, imm, reg) +#define pslld_m2r(var,reg) mmx_m2r (pslld, var, reg) +#define pslld_r2r(regs,regd) mmx_r2r (pslld, regs, regd) +#define psllq_i2r(imm,reg) mmx_i2r (psllq, imm, reg) +#define psllq_m2r(var,reg) mmx_m2r (psllq, var, reg) +#define psllq_r2r(regs,regd) mmx_r2r (psllq, regs, regd) +#define psllw_i2r(imm,reg) mmx_i2r (psllw, imm, reg) +#define psllw_m2r(var,reg) mmx_m2r (psllw, var, reg) +#define psllw_r2r(regs,regd) mmx_r2r (psllw, regs, regd) -#define psrad_i2r(imm,reg) mmx_i2r (psrad, imm, reg) -#define psrad_m2r(var,reg) mmx_m2r (psrad, var, reg) -#define psrad_r2r(regs,regd) mmx_r2r (psrad, regs, regd) -#define psraw_i2r(imm,reg) mmx_i2r (psraw, imm, reg) -#define psraw_m2r(var,reg) mmx_m2r (psraw, var, reg) -#define psraw_r2r(regs,regd) mmx_r2r (psraw, regs, regd) +#define psrad_i2r(imm,reg) mmx_i2r (psrad, imm, reg) +#define psrad_m2r(var,reg) mmx_m2r (psrad, var, reg) +#define psrad_r2r(regs,regd) mmx_r2r (psrad, regs, regd) +#define psraw_i2r(imm,reg) mmx_i2r (psraw, imm, reg) +#define psraw_m2r(var,reg) mmx_m2r (psraw, var, reg) +#define psraw_r2r(regs,regd) mmx_r2r (psraw, regs, regd) -#define psrld_i2r(imm,reg) mmx_i2r (psrld, imm, reg) -#define psrld_m2r(var,reg) mmx_m2r (psrld, var, reg) -#define psrld_r2r(regs,regd) mmx_r2r (psrld, regs, regd) -#define psrlq_i2r(imm,reg) mmx_i2r (psrlq, imm, reg) -#define psrlq_m2r(var,reg) mmx_m2r (psrlq, var, reg) -#define psrlq_r2r(regs,regd) mmx_r2r (psrlq, regs, regd) -#define psrlw_i2r(imm,reg) mmx_i2r (psrlw, imm, reg) -#define psrlw_m2r(var,reg) mmx_m2r (psrlw, var, reg) -#define psrlw_r2r(regs,regd) mmx_r2r (psrlw, regs, regd) +#define psrld_i2r(imm,reg) mmx_i2r (psrld, imm, reg) +#define psrld_m2r(var,reg) mmx_m2r (psrld, var, reg) +#define psrld_r2r(regs,regd) mmx_r2r (psrld, regs, regd) +#define psrlq_i2r(imm,reg) mmx_i2r (psrlq, imm, reg) +#define psrlq_m2r(var,reg) mmx_m2r (psrlq, var, reg) +#define psrlq_r2r(regs,regd) mmx_r2r (psrlq, regs, regd) +#define psrlw_i2r(imm,reg) mmx_i2r (psrlw, imm, reg) +#define psrlw_m2r(var,reg) mmx_m2r (psrlw, var, reg) +#define psrlw_r2r(regs,regd) mmx_r2r (psrlw, regs, regd) -#define psubb_m2r(var,reg) mmx_m2r (psubb, var, reg) -#define psubb_r2r(regs,regd) mmx_r2r (psubb, regs, regd) -#define psubd_m2r(var,reg) mmx_m2r (psubd, var, reg) -#define psubd_r2r(regs,regd) mmx_r2r (psubd, regs, regd) -#define psubw_m2r(var,reg) mmx_m2r (psubw, var, reg) -#define psubw_r2r(regs,regd) mmx_r2r (psubw, regs, regd) +#define psubb_m2r(var,reg) mmx_m2r (psubb, var, reg) +#define psubb_r2r(regs,regd) mmx_r2r (psubb, regs, regd) +#define psubd_m2r(var,reg) mmx_m2r (psubd, var, reg) +#define psubd_r2r(regs,regd) mmx_r2r (psubd, regs, regd) +#define psubw_m2r(var,reg) mmx_m2r (psubw, var, reg) +#define psubw_r2r(regs,regd) mmx_r2r (psubw, regs, regd) -#define psubsb_m2r(var,reg) mmx_m2r (psubsb, var, reg) -#define psubsb_r2r(regs,regd) mmx_r2r (psubsb, regs, regd) -#define psubsw_m2r(var,reg) mmx_m2r (psubsw, var, reg) -#define psubsw_r2r(regs,regd) mmx_r2r (psubsw, regs, regd) +#define psubsb_m2r(var,reg) mmx_m2r (psubsb, var, reg) +#define psubsb_r2r(regs,regd) mmx_r2r (psubsb, regs, regd) +#define psubsw_m2r(var,reg) mmx_m2r (psubsw, var, reg) +#define psubsw_r2r(regs,regd) mmx_r2r (psubsw, regs, regd) -#define psubusb_m2r(var,reg) mmx_m2r (psubusb, var, reg) -#define psubusb_r2r(regs,regd) mmx_r2r (psubusb, regs, regd) -#define psubusw_m2r(var,reg) mmx_m2r (psubusw, var, reg) -#define psubusw_r2r(regs,regd) mmx_r2r (psubusw, regs, regd) +#define psubusb_m2r(var,reg) mmx_m2r (psubusb, var, reg) +#define psubusb_r2r(regs,regd) mmx_r2r (psubusb, regs, regd) +#define psubusw_m2r(var,reg) mmx_m2r (psubusw, var, reg) +#define psubusw_r2r(regs,regd) mmx_r2r (psubusw, regs, regd) -#define punpckhbw_m2r(var,reg) mmx_m2r (punpckhbw, var, reg) -#define punpckhbw_r2r(regs,regd) mmx_r2r (punpckhbw, regs, regd) -#define punpckhdq_m2r(var,reg) mmx_m2r (punpckhdq, var, reg) -#define punpckhdq_r2r(regs,regd) mmx_r2r (punpckhdq, regs, regd) -#define punpckhwd_m2r(var,reg) mmx_m2r (punpckhwd, var, reg) -#define punpckhwd_r2r(regs,regd) mmx_r2r (punpckhwd, regs, regd) +#define punpckhbw_m2r(var,reg) mmx_m2r (punpckhbw, var, reg) +#define punpckhbw_r2r(regs,regd) mmx_r2r (punpckhbw, regs, regd) +#define punpckhdq_m2r(var,reg) mmx_m2r (punpckhdq, var, reg) +#define punpckhdq_r2r(regs,regd) mmx_r2r (punpckhdq, regs, regd) +#define punpckhwd_m2r(var,reg) mmx_m2r (punpckhwd, var, reg) +#define punpckhwd_r2r(regs,regd) mmx_r2r (punpckhwd, regs, regd) -#define punpcklbw_m2r(var,reg) mmx_m2r (punpcklbw, var, reg) -#define punpcklbw_r2r(regs,regd) mmx_r2r (punpcklbw, regs, regd) -#define punpckldq_m2r(var,reg) mmx_m2r (punpckldq, var, reg) -#define punpckldq_r2r(regs,regd) mmx_r2r (punpckldq, regs, regd) -#define punpcklwd_m2r(var,reg) mmx_m2r (punpcklwd, var, reg) -#define punpcklwd_r2r(regs,regd) mmx_r2r (punpcklwd, regs, regd) +#define punpcklbw_m2r(var,reg) mmx_m2r (punpcklbw, var, reg) +#define punpcklbw_r2r(regs,regd) mmx_r2r (punpcklbw, regs, regd) +#define punpckldq_m2r(var,reg) mmx_m2r (punpckldq, var, reg) +#define punpckldq_r2r(regs,regd) mmx_r2r (punpckldq, regs, regd) +#define punpcklwd_m2r(var,reg) mmx_m2r (punpcklwd, var, reg) +#define punpcklwd_r2r(regs,regd) mmx_r2r (punpcklwd, regs, regd) -#define pxor_m2r(var,reg) mmx_m2r (pxor, var, reg) -#define pxor_r2r(regs,regd) mmx_r2r (pxor, regs, regd) +#define pxor_m2r(var,reg) mmx_m2r (pxor, var, reg) +#define pxor_r2r(regs,regd) mmx_r2r (pxor, regs, regd) /* 3DNOW extensions */ -#define pavgusb_m2r(var,reg) mmx_m2r (pavgusb, var, reg) -#define pavgusb_r2r(regs,regd) mmx_r2r (pavgusb, regs, regd) +#define pavgusb_m2r(var,reg) mmx_m2r (pavgusb, var, reg) +#define pavgusb_r2r(regs,regd) mmx_r2r (pavgusb, regs, regd) /* AMD MMX extensions - also available in intel SSE */ -#define mmx_m2ri(op,mem,reg,imm) \ +#define mmx_m2ri(op,mem,reg,imm) \ __asm__ __volatile__ (#op " %1, %0, %%" #reg \ : /* nothing */ \ : "X" (mem), "X" (imm)) -#define mmx_r2ri(op,regs,regd,imm) \ +#define mmx_r2ri(op,regs,regd,imm) \ __asm__ __volatile__ (#op " %0, %%" #regs ", %%" #regd \ : /* nothing */ \ : "X" (imm) ) -#define mmx_fetch(mem,hint) \ - __asm__ __volatile__ ("prefetch" #hint " %0" \ - : /* nothing */ \ - : "X" (mem)) +#define mmx_fetch(mem,hint) \ + __asm__ __volatile__ ("prefetch" #hint " %0" \ + : /* nothing */ \ + : "X" (mem)) -#define maskmovq(regs,maskreg) mmx_r2ri (maskmovq, regs, maskreg) +#define maskmovq(regs,maskreg) mmx_r2ri (maskmovq, regs, maskreg) -#define movntq_r2m(mmreg,var) mmx_r2m (movntq, mmreg, var) +#define movntq_r2m(mmreg,var) mmx_r2m (movntq, mmreg, var) -#define pavgb_m2r(var,reg) mmx_m2r (pavgb, var, reg) -#define pavgb_r2r(regs,regd) mmx_r2r (pavgb, regs, regd) -#define pavgw_m2r(var,reg) mmx_m2r (pavgw, var, reg) -#define pavgw_r2r(regs,regd) mmx_r2r (pavgw, regs, regd) +#define pavgb_m2r(var,reg) mmx_m2r (pavgb, var, reg) +#define pavgb_r2r(regs,regd) mmx_r2r (pavgb, regs, regd) +#define pavgw_m2r(var,reg) mmx_m2r (pavgw, var, reg) +#define pavgw_r2r(regs,regd) mmx_r2r (pavgw, regs, regd) -#define pextrw_r2r(mmreg,reg,imm) mmx_r2ri (pextrw, mmreg, reg, imm) +#define pextrw_r2r(mmreg,reg,imm) mmx_r2ri (pextrw, mmreg, reg, imm) -#define pinsrw_r2r(reg,mmreg,imm) mmx_r2ri (pinsrw, reg, mmreg, imm) +#define pinsrw_r2r(reg,mmreg,imm) mmx_r2ri (pinsrw, reg, mmreg, imm) -#define pmaxsw_m2r(var,reg) mmx_m2r (pmaxsw, var, reg) -#define pmaxsw_r2r(regs,regd) mmx_r2r (pmaxsw, regs, regd) +#define pmaxsw_m2r(var,reg) mmx_m2r (pmaxsw, var, reg) +#define pmaxsw_r2r(regs,regd) mmx_r2r (pmaxsw, regs, regd) -#define pmaxub_m2r(var,reg) mmx_m2r (pmaxub, var, reg) -#define pmaxub_r2r(regs,regd) mmx_r2r (pmaxub, regs, regd) +#define pmaxub_m2r(var,reg) mmx_m2r (pmaxub, var, reg) +#define pmaxub_r2r(regs,regd) mmx_r2r (pmaxub, regs, regd) -#define pminsw_m2r(var,reg) mmx_m2r (pminsw, var, reg) -#define pminsw_r2r(regs,regd) mmx_r2r (pminsw, regs, regd) +#define pminsw_m2r(var,reg) mmx_m2r (pminsw, var, reg) +#define pminsw_r2r(regs,regd) mmx_r2r (pminsw, regs, regd) -#define pminub_m2r(var,reg) mmx_m2r (pminub, var, reg) -#define pminub_r2r(regs,regd) mmx_r2r (pminub, regs, regd) +#define pminub_m2r(var,reg) mmx_m2r (pminub, var, reg) +#define pminub_r2r(regs,regd) mmx_r2r (pminub, regs, regd) -#define pmovmskb(mmreg,reg) \ - __asm__ __volatile__ ("movmskps %" #mmreg ", %" #reg) +#define pmovmskb(mmreg,reg) \ + __asm__ __volatile__ ("movmskps %" #mmreg ", %" #reg) -#define pmulhuw_m2r(var,reg) mmx_m2r (pmulhuw, var, reg) -#define pmulhuw_r2r(regs,regd) mmx_r2r (pmulhuw, regs, regd) +#define pmulhuw_m2r(var,reg) mmx_m2r (pmulhuw, var, reg) +#define pmulhuw_r2r(regs,regd) mmx_r2r (pmulhuw, regs, regd) -#define prefetcht0(mem) mmx_fetch (mem, t0) -#define prefetcht1(mem) mmx_fetch (mem, t1) -#define prefetcht2(mem) mmx_fetch (mem, t2) -#define prefetchnta(mem) mmx_fetch (mem, nta) +#define prefetcht0(mem) mmx_fetch (mem, t0) +#define prefetcht1(mem) mmx_fetch (mem, t1) +#define prefetcht2(mem) mmx_fetch (mem, t2) +#define prefetchnta(mem) mmx_fetch (mem, nta) -#define psadbw_m2r(var,reg) mmx_m2r (psadbw, var, reg) -#define psadbw_r2r(regs,regd) mmx_r2r (psadbw, regs, regd) +#define psadbw_m2r(var,reg) mmx_m2r (psadbw, var, reg) +#define psadbw_r2r(regs,regd) mmx_r2r (psadbw, regs, regd) -#define pshufw_m2r(var,reg,imm) mmx_m2ri(pshufw, var, reg, imm) -#define pshufw_r2r(regs,regd,imm) mmx_r2ri(pshufw, regs, regd, imm) +#define pshufw_m2r(var,reg,imm) mmx_m2ri(pshufw, var, reg, imm) +#define pshufw_r2r(regs,regd,imm) mmx_r2ri(pshufw, regs, regd, imm) -#define sfence() __asm__ __volatile__ ("sfence\n\t") +#define sfence() __asm__ __volatile__ ("sfence\n\t") /* SSE2 */ -#define pshufhw_m2r(var,reg,imm) mmx_m2ri(pshufhw, var, reg, imm) -#define pshufhw_r2r(regs,regd,imm) mmx_r2ri(pshufhw, regs, regd, imm) -#define pshuflw_m2r(var,reg,imm) mmx_m2ri(pshuflw, var, reg, imm) -#define pshuflw_r2r(regs,regd,imm) mmx_r2ri(pshuflw, regs, regd, imm) +#define pshufhw_m2r(var,reg,imm) mmx_m2ri(pshufhw, var, reg, imm) +#define pshufhw_r2r(regs,regd,imm) mmx_r2ri(pshufhw, regs, regd, imm) +#define pshuflw_m2r(var,reg,imm) mmx_m2ri(pshuflw, var, reg, imm) +#define pshuflw_r2r(regs,regd,imm) mmx_r2ri(pshuflw, regs, regd, imm) -#define pshufd_r2r(regs,regd,imm) mmx_r2ri(pshufd, regs, regd, imm) +#define pshufd_r2r(regs,regd,imm) mmx_r2ri(pshufd, regs, regd, imm) -#define movdqa_m2r(var,reg) mmx_m2r (movdqa, var, reg) -#define movdqa_r2m(reg,var) mmx_r2m (movdqa, reg, var) -#define movdqa_r2r(regs,regd) mmx_r2r (movdqa, regs, regd) -#define movdqu_m2r(var,reg) mmx_m2r (movdqu, var, reg) -#define movdqu_r2m(reg,var) mmx_r2m (movdqu, reg, var) -#define movdqu_r2r(regs,regd) mmx_r2r (movdqu, regs, regd) +#define movdqa_m2r(var,reg) mmx_m2r (movdqa, var, reg) +#define movdqa_r2m(reg,var) mmx_r2m (movdqa, reg, var) +#define movdqa_r2r(regs,regd) mmx_r2r (movdqa, regs, regd) +#define movdqu_m2r(var,reg) mmx_m2r (movdqu, var, reg) +#define movdqu_r2m(reg,var) mmx_r2m (movdqu, reg, var) +#define movdqu_r2r(regs,regd) mmx_r2r (movdqu, regs, regd) -#define pmullw_r2m(reg,var) mmx_r2m (pmullw, reg, var) +#define pmullw_r2m(reg,var) mmx_r2m (pmullw, reg, var) -#define pslldq_i2r(imm,reg) mmx_i2r (pslldq, imm, reg) -#define psrldq_i2r(imm,reg) mmx_i2r (psrldq, imm, reg) +#define pslldq_i2r(imm,reg) mmx_i2r (pslldq, imm, reg) +#define psrldq_i2r(imm,reg) mmx_i2r (psrldq, imm, reg) -#define punpcklqdq_r2r(regs,regd) mmx_r2r (punpcklqdq, regs, regd) -#define punpckhqdq_r2r(regs,regd) mmx_r2r (punpckhqdq, regs, regd) +#define punpcklqdq_r2r(regs,regd) mmx_r2r (punpcklqdq, regs, regd) +#define punpckhqdq_r2r(regs,regd) mmx_r2r (punpckhqdq, regs, regd) #endif /* AVCODEC_I386MMX_H */
--- a/i386/motion_est_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/motion_est_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -34,33 +34,33 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm2 \n\t" - "movq (%2, %%"REG_a"), %%mm4 \n\t" - "add %3, %%"REG_a" \n\t" - "psubusb %%mm0, %%mm2 \n\t" - "psubusb %%mm4, %%mm0 \n\t" - "movq (%1, %%"REG_a"), %%mm1 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "movq (%2, %%"REG_a"), %%mm5 \n\t" - "psubusb %%mm1, %%mm3 \n\t" - "psubusb %%mm5, %%mm1 \n\t" - "por %%mm2, %%mm0 \n\t" - "por %%mm1, %%mm3 \n\t" - "movq %%mm0, %%mm1 \n\t" - "movq %%mm3, %%mm2 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm3 \n\t" - "punpckhbw %%mm7, %%mm2 \n\t" - "paddw %%mm1, %%mm0 \n\t" - "paddw %%mm3, %%mm2 \n\t" - "paddw %%mm2, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %3, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm2 \n\t" + "movq (%2, %%"REG_a"), %%mm4 \n\t" + "add %3, %%"REG_a" \n\t" + "psubusb %%mm0, %%mm2 \n\t" + "psubusb %%mm4, %%mm0 \n\t" + "movq (%1, %%"REG_a"), %%mm1 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "movq (%2, %%"REG_a"), %%mm5 \n\t" + "psubusb %%mm1, %%mm3 \n\t" + "psubusb %%mm5, %%mm1 \n\t" + "por %%mm2, %%mm0 \n\t" + "por %%mm1, %%mm3 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm3, %%mm2 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm3 \n\t" + "punpckhbw %%mm7, %%mm2 \n\t" + "paddw %%mm1, %%mm0 \n\t" + "paddw %%mm3, %%mm2 \n\t" + "paddw %%mm2, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %3, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1 - len), "r" (blk2 - len), "r" ((long)stride) ); @@ -70,19 +70,19 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm2 \n\t" - "psadbw %%mm2, %%mm0 \n\t" - "add %3, %%"REG_a" \n\t" - "movq (%1, %%"REG_a"), %%mm1 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "psadbw %%mm1, %%mm3 \n\t" - "paddw %%mm3, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %3, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm2 \n\t" + "psadbw %%mm2, %%mm0 \n\t" + "add %3, %%"REG_a" \n\t" + "movq (%1, %%"REG_a"), %%mm1 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "psadbw %%mm1, %%mm3 \n\t" + "paddw %%mm3, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %3, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1 - len), "r" (blk2 - len), "r" ((long)stride) ); @@ -92,23 +92,23 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm2 \n\t" - "pavgb %%mm2, %%mm0 \n\t" - "movq (%3, %%"REG_a"), %%mm2 \n\t" - "psadbw %%mm2, %%mm0 \n\t" - "add %4, %%"REG_a" \n\t" - "movq (%1, %%"REG_a"), %%mm1 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "pavgb %%mm1, %%mm3 \n\t" - "movq (%3, %%"REG_a"), %%mm1 \n\t" - "psadbw %%mm1, %%mm3 \n\t" - "paddw %%mm3, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %4, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm2 \n\t" + "pavgb %%mm2, %%mm0 \n\t" + "movq (%3, %%"REG_a"), %%mm2 \n\t" + "psadbw %%mm2, %%mm0 \n\t" + "add %4, %%"REG_a" \n\t" + "movq (%1, %%"REG_a"), %%mm1 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "pavgb %%mm1, %%mm3 \n\t" + "movq (%3, %%"REG_a"), %%mm1 \n\t" + "psadbw %%mm1, %%mm3 \n\t" + "paddw %%mm3, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %4, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1a - len), "r" (blk1b -len), "r" (blk2 - len), "r" ((long)stride) ); @@ -118,34 +118,34 @@ { //FIXME reuse src long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "movq "MANGLE(bone)", %%mm5 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm2 \n\t" - "movq 1(%1, %%"REG_a"), %%mm1 \n\t" - "movq 1(%2, %%"REG_a"), %%mm3 \n\t" - "pavgb %%mm2, %%mm0 \n\t" - "pavgb %%mm1, %%mm3 \n\t" - "psubusb %%mm5, %%mm3 \n\t" - "pavgb %%mm3, %%mm0 \n\t" - "movq (%3, %%"REG_a"), %%mm2 \n\t" - "psadbw %%mm2, %%mm0 \n\t" - "add %4, %%"REG_a" \n\t" - "movq (%1, %%"REG_a"), %%mm1 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "movq 1(%1, %%"REG_a"), %%mm2 \n\t" - "movq 1(%2, %%"REG_a"), %%mm4 \n\t" - "pavgb %%mm3, %%mm1 \n\t" - "pavgb %%mm4, %%mm2 \n\t" - "psubusb %%mm5, %%mm2 \n\t" - "pavgb %%mm1, %%mm2 \n\t" - "movq (%3, %%"REG_a"), %%mm1 \n\t" - "psadbw %%mm1, %%mm2 \n\t" - "paddw %%mm2, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %4, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "movq "MANGLE(bone)", %%mm5 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm2 \n\t" + "movq 1(%1, %%"REG_a"), %%mm1 \n\t" + "movq 1(%2, %%"REG_a"), %%mm3 \n\t" + "pavgb %%mm2, %%mm0 \n\t" + "pavgb %%mm1, %%mm3 \n\t" + "psubusb %%mm5, %%mm3 \n\t" + "pavgb %%mm3, %%mm0 \n\t" + "movq (%3, %%"REG_a"), %%mm2 \n\t" + "psadbw %%mm2, %%mm0 \n\t" + "add %4, %%"REG_a" \n\t" + "movq (%1, %%"REG_a"), %%mm1 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "movq 1(%1, %%"REG_a"), %%mm2 \n\t" + "movq 1(%2, %%"REG_a"), %%mm4 \n\t" + "pavgb %%mm3, %%mm1 \n\t" + "pavgb %%mm4, %%mm2 \n\t" + "psubusb %%mm5, %%mm2 \n\t" + "pavgb %%mm1, %%mm2 \n\t" + "movq (%3, %%"REG_a"), %%mm1 \n\t" + "psadbw %%mm1, %%mm2 \n\t" + "paddw %%mm2, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %4, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1 - len), "r" (blk1 - len + stride), "r" (blk2 - len), "r" ((long)stride) ); @@ -155,35 +155,35 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm2 \n\t" - "movq (%2, %%"REG_a"), %%mm3 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "paddw %%mm0, %%mm1 \n\t" - "paddw %%mm2, %%mm3 \n\t" - "movq (%3, %%"REG_a"), %%mm4 \n\t" - "movq (%3, %%"REG_a"), %%mm2 \n\t" - "paddw %%mm5, %%mm1 \n\t" - "paddw %%mm5, %%mm3 \n\t" - "psrlw $1, %%mm1 \n\t" - "psrlw $1, %%mm3 \n\t" - "packuswb %%mm3, %%mm1 \n\t" - "psubusb %%mm1, %%mm4 \n\t" - "psubusb %%mm2, %%mm1 \n\t" - "por %%mm4, %%mm1 \n\t" - "movq %%mm1, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "paddw %%mm1, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %4, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm2 \n\t" + "movq (%2, %%"REG_a"), %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "paddw %%mm0, %%mm1 \n\t" + "paddw %%mm2, %%mm3 \n\t" + "movq (%3, %%"REG_a"), %%mm4 \n\t" + "movq (%3, %%"REG_a"), %%mm2 \n\t" + "paddw %%mm5, %%mm1 \n\t" + "paddw %%mm5, %%mm3 \n\t" + "psrlw $1, %%mm1 \n\t" + "psrlw $1, %%mm3 \n\t" + "packuswb %%mm3, %%mm1 \n\t" + "psubusb %%mm1, %%mm4 \n\t" + "psubusb %%mm2, %%mm1 \n\t" + "por %%mm4, %%mm1 \n\t" + "movq %%mm1, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "paddw %%mm1, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %4, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1a - len), "r" (blk1b -len), "r" (blk2 - len), "r" ((long)stride) ); @@ -193,47 +193,47 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" - "1: \n\t" - "movq (%1, %%"REG_a"), %%mm0 \n\t" - "movq (%2, %%"REG_a"), %%mm1 \n\t" - "movq %%mm0, %%mm4 \n\t" - "movq %%mm1, %%mm2 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm1 \n\t" - "punpckhbw %%mm7, %%mm4 \n\t" - "punpckhbw %%mm7, %%mm2 \n\t" - "paddw %%mm1, %%mm0 \n\t" - "paddw %%mm2, %%mm4 \n\t" - "movq 1(%1, %%"REG_a"), %%mm2 \n\t" - "movq 1(%2, %%"REG_a"), %%mm3 \n\t" - "movq %%mm2, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm1 \n\t" - "paddw %%mm0, %%mm2 \n\t" - "paddw %%mm4, %%mm1 \n\t" - "movq %%mm3, %%mm4 \n\t" - "punpcklbw %%mm7, %%mm3 \n\t" - "punpckhbw %%mm7, %%mm4 \n\t" - "paddw %%mm3, %%mm2 \n\t" - "paddw %%mm4, %%mm1 \n\t" - "movq (%3, %%"REG_a"), %%mm3 \n\t" - "movq (%3, %%"REG_a"), %%mm4 \n\t" - "paddw %%mm5, %%mm2 \n\t" - "paddw %%mm5, %%mm1 \n\t" - "psrlw $2, %%mm2 \n\t" - "psrlw $2, %%mm1 \n\t" - "packuswb %%mm1, %%mm2 \n\t" - "psubusb %%mm2, %%mm3 \n\t" - "psubusb %%mm4, %%mm2 \n\t" - "por %%mm3, %%mm2 \n\t" - "movq %%mm2, %%mm0 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" - "punpckhbw %%mm7, %%mm2 \n\t" - "paddw %%mm2, %%mm0 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "add %4, %%"REG_a" \n\t" - " js 1b \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%1, %%"REG_a"), %%mm0 \n\t" + "movq (%2, %%"REG_a"), %%mm1 \n\t" + "movq %%mm0, %%mm4 \n\t" + "movq %%mm1, %%mm2 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm1 \n\t" + "punpckhbw %%mm7, %%mm4 \n\t" + "punpckhbw %%mm7, %%mm2 \n\t" + "paddw %%mm1, %%mm0 \n\t" + "paddw %%mm2, %%mm4 \n\t" + "movq 1(%1, %%"REG_a"), %%mm2 \n\t" + "movq 1(%2, %%"REG_a"), %%mm3 \n\t" + "movq %%mm2, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "paddw %%mm0, %%mm2 \n\t" + "paddw %%mm4, %%mm1 \n\t" + "movq %%mm3, %%mm4 \n\t" + "punpcklbw %%mm7, %%mm3 \n\t" + "punpckhbw %%mm7, %%mm4 \n\t" + "paddw %%mm3, %%mm2 \n\t" + "paddw %%mm4, %%mm1 \n\t" + "movq (%3, %%"REG_a"), %%mm3 \n\t" + "movq (%3, %%"REG_a"), %%mm4 \n\t" + "paddw %%mm5, %%mm2 \n\t" + "paddw %%mm5, %%mm1 \n\t" + "psrlw $2, %%mm2 \n\t" + "psrlw $2, %%mm1 \n\t" + "packuswb %%mm1, %%mm2 \n\t" + "psubusb %%mm2, %%mm3 \n\t" + "psubusb %%mm4, %%mm2 \n\t" + "por %%mm3, %%mm2 \n\t" + "movq %%mm2, %%mm0 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm2 \n\t" + "paddw %%mm2, %%mm0 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "add %4, %%"REG_a" \n\t" + " js 1b \n\t" : "+a" (len) : "r" (blk1 - len), "r" (blk1 -len + stride), "r" (blk2 - len), "r" ((long)stride) ); @@ -243,13 +243,13 @@ { int ret; asm volatile( - "movq %%mm6, %%mm0 \n\t" - "psrlq $32, %%mm6 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "movq %%mm6, %%mm0 \n\t" - "psrlq $16, %%mm6 \n\t" - "paddw %%mm0, %%mm6 \n\t" - "movd %%mm6, %0 \n\t" + "movq %%mm6, %%mm0 \n\t" + "psrlq $32, %%mm6 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "movq %%mm6, %%mm0 \n\t" + "psrlq $16, %%mm6 \n\t" + "paddw %%mm0, %%mm6 \n\t" + "movd %%mm6, %0 \n\t" : "=r" (ret) ); return ret&0xFFFF; @@ -259,7 +259,7 @@ { int ret; asm volatile( - "movd %%mm6, %0 \n\t" + "movd %%mm6, %0 \n\t" : "=r" (ret) ); return ret; @@ -270,8 +270,8 @@ static int sad8_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ assert(h==8);\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t":);\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t":);\ \ sad8_1_ ## suf(blk1, blk2, stride, 8);\ \ @@ -280,9 +280,9 @@ static int sad8_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ assert(h==8);\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[1]) \ );\ \ @@ -294,9 +294,9 @@ static int sad8_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ assert(h==8);\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[1]) \ );\ \ @@ -308,9 +308,9 @@ static int sad8_xy2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ assert(h==8);\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[2]) \ );\ \ @@ -321,8 +321,8 @@ \ static int sad16_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t":);\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t":);\ \ sad8_1_ ## suf(blk1 , blk2 , stride, h);\ sad8_1_ ## suf(blk1+8, blk2+8, stride, h);\ @@ -331,9 +331,9 @@ }\ static int sad16_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[1]) \ );\ \ @@ -344,9 +344,9 @@ }\ static int sad16_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[1]) \ );\ \ @@ -357,9 +357,9 @@ }\ static int sad16_xy2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\ {\ - asm volatile("pxor %%mm7, %%mm7 \n\t"\ - "pxor %%mm6, %%mm6 \n\t"\ - "movq %0, %%mm5 \n\t"\ + asm volatile("pxor %%mm7, %%mm7 \n\t"\ + "pxor %%mm6, %%mm6 \n\t"\ + "movq %0, %%mm5 \n\t"\ :: "m"(round_tab[2]) \ );\ \ @@ -384,15 +384,15 @@ c->pix_abs[1][2] = sad8_y2_mmx; c->pix_abs[1][3] = sad8_xy2_mmx; - c->sad[0]= sad16_mmx; + c->sad[0]= sad16_mmx; c->sad[1]= sad8_mmx; } if (mm_flags & MM_MMXEXT) { - c->pix_abs[0][0] = sad16_mmx2; - c->pix_abs[1][0] = sad8_mmx2; + c->pix_abs[0][0] = sad16_mmx2; + c->pix_abs[1][0] = sad8_mmx2; - c->sad[0]= sad16_mmx2; - c->sad[1]= sad8_mmx2; + c->sad[0]= sad16_mmx2; + c->sad[1]= sad8_mmx2; if(!(avctx->flags & CODEC_FLAG_BITEXACT)){ c->pix_abs[0][1] = sad16_x2_mmx2;
--- a/i386/mpegvideo_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/mpegvideo_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -57,52 +57,52 @@ nCoeffs= s->inter_scantable.raster_end[ s->block_last_index[n] ]; //printf("%d %d ", qmul, qadd); asm volatile( - "movd %1, %%mm6 \n\t" //qmul - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "movd %2, %%mm5 \n\t" //qadd - "pxor %%mm7, %%mm7 \n\t" - "packssdw %%mm5, %%mm5 \n\t" - "packssdw %%mm5, %%mm5 \n\t" - "psubw %%mm5, %%mm7 \n\t" - "pxor %%mm4, %%mm4 \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %3), %%mm0 \n\t" - "movq 8(%0, %3), %%mm1 \n\t" + "movd %1, %%mm6 \n\t" //qmul + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "movd %2, %%mm5 \n\t" //qadd + "pxor %%mm7, %%mm7 \n\t" + "packssdw %%mm5, %%mm5 \n\t" + "packssdw %%mm5, %%mm5 \n\t" + "psubw %%mm5, %%mm7 \n\t" + "pxor %%mm4, %%mm4 \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %3), %%mm0 \n\t" + "movq 8(%0, %3), %%mm1 \n\t" - "pmullw %%mm6, %%mm0 \n\t" - "pmullw %%mm6, %%mm1 \n\t" + "pmullw %%mm6, %%mm0 \n\t" + "pmullw %%mm6, %%mm1 \n\t" - "movq (%0, %3), %%mm2 \n\t" - "movq 8(%0, %3), %%mm3 \n\t" + "movq (%0, %3), %%mm2 \n\t" + "movq 8(%0, %3), %%mm3 \n\t" - "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" - "paddw %%mm7, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" + "paddw %%mm7, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" - "pxor %%mm0, %%mm2 \n\t" - "pxor %%mm1, %%mm3 \n\t" + "pxor %%mm0, %%mm2 \n\t" + "pxor %%mm1, %%mm3 \n\t" - "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0 - "pandn %%mm2, %%mm0 \n\t" - "pandn %%mm3, %%mm1 \n\t" + "pandn %%mm2, %%mm0 \n\t" + "pandn %%mm3, %%mm1 \n\t" - "movq %%mm0, (%0, %3) \n\t" - "movq %%mm1, 8(%0, %3) \n\t" + "movq %%mm0, (%0, %3) \n\t" + "movq %%mm1, 8(%0, %3) \n\t" - "add $16, %3 \n\t" - "jng 1b \n\t" - ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs)) - : "memory" - ); + "add $16, %3 \n\t" + "jng 1b \n\t" + ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs)) + : "memory" + ); block[0]= level; } @@ -120,52 +120,52 @@ nCoeffs= s->inter_scantable.raster_end[ s->block_last_index[n] ]; //printf("%d %d ", qmul, qadd); asm volatile( - "movd %1, %%mm6 \n\t" //qmul - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "movd %2, %%mm5 \n\t" //qadd - "pxor %%mm7, %%mm7 \n\t" - "packssdw %%mm5, %%mm5 \n\t" - "packssdw %%mm5, %%mm5 \n\t" - "psubw %%mm5, %%mm7 \n\t" - "pxor %%mm4, %%mm4 \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %3), %%mm0 \n\t" - "movq 8(%0, %3), %%mm1 \n\t" + "movd %1, %%mm6 \n\t" //qmul + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "movd %2, %%mm5 \n\t" //qadd + "pxor %%mm7, %%mm7 \n\t" + "packssdw %%mm5, %%mm5 \n\t" + "packssdw %%mm5, %%mm5 \n\t" + "psubw %%mm5, %%mm7 \n\t" + "pxor %%mm4, %%mm4 \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %3), %%mm0 \n\t" + "movq 8(%0, %3), %%mm1 \n\t" - "pmullw %%mm6, %%mm0 \n\t" - "pmullw %%mm6, %%mm1 \n\t" + "pmullw %%mm6, %%mm0 \n\t" + "pmullw %%mm6, %%mm1 \n\t" - "movq (%0, %3), %%mm2 \n\t" - "movq 8(%0, %3), %%mm3 \n\t" + "movq (%0, %3), %%mm2 \n\t" + "movq 8(%0, %3), %%mm3 \n\t" - "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" - "paddw %%mm7, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" + "paddw %%mm7, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" - "pxor %%mm0, %%mm2 \n\t" - "pxor %%mm1, %%mm3 \n\t" + "pxor %%mm0, %%mm2 \n\t" + "pxor %%mm1, %%mm3 \n\t" - "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0 - "pandn %%mm2, %%mm0 \n\t" - "pandn %%mm3, %%mm1 \n\t" + "pandn %%mm2, %%mm0 \n\t" + "pandn %%mm3, %%mm1 \n\t" - "movq %%mm0, (%0, %3) \n\t" - "movq %%mm1, 8(%0, %3) \n\t" + "movq %%mm0, (%0, %3) \n\t" + "movq %%mm1, 8(%0, %3) \n\t" - "add $16, %3 \n\t" - "jng 1b \n\t" - ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs)) - : "memory" - ); + "add $16, %3 \n\t" + "jng 1b \n\t" + ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs)) + : "memory" + ); } @@ -216,54 +216,54 @@ /* XXX: only mpeg1 */ quant_matrix = s->intra_matrix; asm volatile( - "pcmpeqw %%mm7, %%mm7 \n\t" - "psrlw $15, %%mm7 \n\t" - "movd %2, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "mov %3, %%"REG_a" \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %%"REG_a"), %%mm0 \n\t" - "movq 8(%0, %%"REG_a"), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm4 \n\t" - "movq 8(%1, %%"REG_a"), %%mm5 \n\t" - "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] - "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) - "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) - "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q - "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q - "pxor %%mm4, %%mm4 \n\t" - "pxor %%mm5, %%mm5 \n\t" // FIXME slow - "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 - "psraw $3, %%mm0 \n\t" - "psraw $3, %%mm1 \n\t" - "psubw %%mm7, %%mm0 \n\t" - "psubw %%mm7, %%mm1 \n\t" - "por %%mm7, %%mm0 \n\t" - "por %%mm7, %%mm1 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" - "psubw %%mm3, %%mm1 \n\t" - "pandn %%mm0, %%mm4 \n\t" - "pandn %%mm1, %%mm5 \n\t" - "movq %%mm4, (%0, %%"REG_a") \n\t" - "movq %%mm5, 8(%0, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm7 \n\t" + "psrlw $15, %%mm7 \n\t" + "movd %2, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "mov %3, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %%"REG_a"), %%mm0 \n\t" + "movq 8(%0, %%"REG_a"), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm4 \n\t" + "movq 8(%1, %%"REG_a"), %%mm5 \n\t" + "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] + "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) + "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) + "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q + "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q + "pxor %%mm4, %%mm4 \n\t" + "pxor %%mm5, %%mm5 \n\t" // FIXME slow + "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 + "psraw $3, %%mm0 \n\t" + "psraw $3, %%mm1 \n\t" + "psubw %%mm7, %%mm0 \n\t" + "psubw %%mm7, %%mm1 \n\t" + "por %%mm7, %%mm0 \n\t" + "por %%mm7, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "pandn %%mm0, %%mm4 \n\t" + "pandn %%mm1, %%mm5 \n\t" + "movq %%mm4, (%0, %%"REG_a") \n\t" + "movq %%mm5, 8(%0, %%"REG_a") \n\t" - "add $16, %%"REG_a" \n\t" - "js 1b \n\t" - ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) - : "%"REG_a, "memory" - ); + "add $16, %%"REG_a" \n\t" + "js 1b \n\t" + ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) + : "%"REG_a, "memory" + ); block[0]= block0; } @@ -279,58 +279,58 @@ quant_matrix = s->inter_matrix; asm volatile( - "pcmpeqw %%mm7, %%mm7 \n\t" - "psrlw $15, %%mm7 \n\t" - "movd %2, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "mov %3, %%"REG_a" \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %%"REG_a"), %%mm0 \n\t" - "movq 8(%0, %%"REG_a"), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm4 \n\t" - "movq 8(%1, %%"REG_a"), %%mm5 \n\t" - "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] - "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) - "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) - "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2 - "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2 - "paddw %%mm7, %%mm0 \n\t" // abs(block[i])*2 + 1 - "paddw %%mm7, %%mm1 \n\t" // abs(block[i])*2 + 1 - "pmullw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q - "pmullw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q - "pxor %%mm4, %%mm4 \n\t" - "pxor %%mm5, %%mm5 \n\t" // FIXME slow - "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 - "psraw $4, %%mm0 \n\t" - "psraw $4, %%mm1 \n\t" - "psubw %%mm7, %%mm0 \n\t" - "psubw %%mm7, %%mm1 \n\t" - "por %%mm7, %%mm0 \n\t" - "por %%mm7, %%mm1 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" - "psubw %%mm3, %%mm1 \n\t" - "pandn %%mm0, %%mm4 \n\t" - "pandn %%mm1, %%mm5 \n\t" - "movq %%mm4, (%0, %%"REG_a") \n\t" - "movq %%mm5, 8(%0, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm7 \n\t" + "psrlw $15, %%mm7 \n\t" + "movd %2, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "mov %3, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %%"REG_a"), %%mm0 \n\t" + "movq 8(%0, %%"REG_a"), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm4 \n\t" + "movq 8(%1, %%"REG_a"), %%mm5 \n\t" + "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] + "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) + "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) + "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2 + "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2 + "paddw %%mm7, %%mm0 \n\t" // abs(block[i])*2 + 1 + "paddw %%mm7, %%mm1 \n\t" // abs(block[i])*2 + 1 + "pmullw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q + "pmullw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q + "pxor %%mm4, %%mm4 \n\t" + "pxor %%mm5, %%mm5 \n\t" // FIXME slow + "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 + "psraw $4, %%mm0 \n\t" + "psraw $4, %%mm1 \n\t" + "psubw %%mm7, %%mm0 \n\t" + "psubw %%mm7, %%mm1 \n\t" + "por %%mm7, %%mm0 \n\t" + "por %%mm7, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "pandn %%mm0, %%mm4 \n\t" + "pandn %%mm1, %%mm5 \n\t" + "movq %%mm4, (%0, %%"REG_a") \n\t" + "movq %%mm5, 8(%0, %%"REG_a") \n\t" - "add $16, %%"REG_a" \n\t" - "js 1b \n\t" - ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) - : "%"REG_a, "memory" - ); + "add $16, %%"REG_a" \n\t" + "js 1b \n\t" + ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) + : "%"REG_a, "memory" + ); } static void dct_unquantize_mpeg2_intra_mmx(MpegEncContext *s, @@ -351,50 +351,50 @@ block0 = block[0] * s->c_dc_scale; quant_matrix = s->intra_matrix; asm volatile( - "pcmpeqw %%mm7, %%mm7 \n\t" - "psrlw $15, %%mm7 \n\t" - "movd %2, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "mov %3, %%"REG_a" \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %%"REG_a"), %%mm0 \n\t" - "movq 8(%0, %%"REG_a"), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm4 \n\t" - "movq 8(%1, %%"REG_a"), %%mm5 \n\t" - "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] - "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) - "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) - "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q - "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q - "pxor %%mm4, %%mm4 \n\t" - "pxor %%mm5, %%mm5 \n\t" // FIXME slow - "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 - "psraw $3, %%mm0 \n\t" - "psraw $3, %%mm1 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" - "psubw %%mm3, %%mm1 \n\t" - "pandn %%mm0, %%mm4 \n\t" - "pandn %%mm1, %%mm5 \n\t" - "movq %%mm4, (%0, %%"REG_a") \n\t" - "movq %%mm5, 8(%0, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm7 \n\t" + "psrlw $15, %%mm7 \n\t" + "movd %2, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "mov %3, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %%"REG_a"), %%mm0 \n\t" + "movq 8(%0, %%"REG_a"), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm4 \n\t" + "movq 8(%1, %%"REG_a"), %%mm5 \n\t" + "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] + "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) + "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) + "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q + "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q + "pxor %%mm4, %%mm4 \n\t" + "pxor %%mm5, %%mm5 \n\t" // FIXME slow + "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 + "psraw $3, %%mm0 \n\t" + "psraw $3, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "pandn %%mm0, %%mm4 \n\t" + "pandn %%mm1, %%mm5 \n\t" + "movq %%mm4, (%0, %%"REG_a") \n\t" + "movq %%mm5, 8(%0, %%"REG_a") \n\t" - "add $16, %%"REG_a" \n\t" - "jng 1b \n\t" - ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) - : "%"REG_a, "memory" - ); + "add $16, %%"REG_a" \n\t" + "jng 1b \n\t" + ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs) + : "%"REG_a, "memory" + ); block[0]= block0; //Note, we dont do mismatch control for intra as errors cannot accumulate } @@ -412,68 +412,68 @@ quant_matrix = s->inter_matrix; asm volatile( - "pcmpeqw %%mm7, %%mm7 \n\t" - "psrlq $48, %%mm7 \n\t" - "movd %2, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "packssdw %%mm6, %%mm6 \n\t" - "mov %3, %%"REG_a" \n\t" - ".balign 16\n\t" - "1: \n\t" - "movq (%0, %%"REG_a"), %%mm0 \n\t" - "movq 8(%0, %%"REG_a"), %%mm1 \n\t" - "movq (%1, %%"REG_a"), %%mm4 \n\t" - "movq 8(%1, %%"REG_a"), %%mm5 \n\t" - "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] - "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 - "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) - "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) - "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2 - "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2 - "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*2*q - "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*2*q - "paddw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q - "paddw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q - "pxor %%mm4, %%mm4 \n\t" - "pxor %%mm5, %%mm5 \n\t" // FIXME slow - "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 - "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 - "psrlw $4, %%mm0 \n\t" - "psrlw $4, %%mm1 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" - "psubw %%mm3, %%mm1 \n\t" - "pandn %%mm0, %%mm4 \n\t" - "pandn %%mm1, %%mm5 \n\t" - "pxor %%mm4, %%mm7 \n\t" - "pxor %%mm5, %%mm7 \n\t" - "movq %%mm4, (%0, %%"REG_a") \n\t" - "movq %%mm5, 8(%0, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm7 \n\t" + "psrlq $48, %%mm7 \n\t" + "movd %2, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "packssdw %%mm6, %%mm6 \n\t" + "mov %3, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "movq (%0, %%"REG_a"), %%mm0 \n\t" + "movq 8(%0, %%"REG_a"), %%mm1 \n\t" + "movq (%1, %%"REG_a"), %%mm4 \n\t" + "movq 8(%1, %%"REG_a"), %%mm5 \n\t" + "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i] + "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i] + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0 + "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0 + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // abs(block[i]) + "psubw %%mm3, %%mm1 \n\t" // abs(block[i]) + "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2 + "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2 + "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*2*q + "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*2*q + "paddw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q + "paddw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q + "pxor %%mm4, %%mm4 \n\t" + "pxor %%mm5, %%mm5 \n\t" // FIXME slow + "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0 + "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0 + "psrlw $4, %%mm0 \n\t" + "psrlw $4, %%mm1 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "pandn %%mm0, %%mm4 \n\t" + "pandn %%mm1, %%mm5 \n\t" + "pxor %%mm4, %%mm7 \n\t" + "pxor %%mm5, %%mm7 \n\t" + "movq %%mm4, (%0, %%"REG_a") \n\t" + "movq %%mm5, 8(%0, %%"REG_a") \n\t" - "add $16, %%"REG_a" \n\t" - "jng 1b \n\t" - "movd 124(%0, %3), %%mm0 \n\t" - "movq %%mm7, %%mm6 \n\t" - "psrlq $32, %%mm7 \n\t" - "pxor %%mm6, %%mm7 \n\t" - "movq %%mm7, %%mm6 \n\t" - "psrlq $16, %%mm7 \n\t" - "pxor %%mm6, %%mm7 \n\t" - "pslld $31, %%mm7 \n\t" - "psrlq $15, %%mm7 \n\t" - "pxor %%mm7, %%mm0 \n\t" - "movd %%mm0, 124(%0, %3) \n\t" + "add $16, %%"REG_a" \n\t" + "jng 1b \n\t" + "movd 124(%0, %3), %%mm0 \n\t" + "movq %%mm7, %%mm6 \n\t" + "psrlq $32, %%mm7 \n\t" + "pxor %%mm6, %%mm7 \n\t" + "movq %%mm7, %%mm6 \n\t" + "psrlq $16, %%mm7 \n\t" + "pxor %%mm6, %%mm7 \n\t" + "pslld $31, %%mm7 \n\t" + "psrlq $15, %%mm7 \n\t" + "pxor %%mm7, %%mm0 \n\t" + "movd %%mm0, 124(%0, %3) \n\t" - ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs) - : "%"REG_a, "memory" - ); + ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs) + : "%"REG_a, "memory" + ); } /* draw the edges of width 'w' of an image of size width, height @@ -488,79 +488,79 @@ ptr = buf; if(w==8) { - asm volatile( - "1: \n\t" - "movd (%0), %%mm0 \n\t" - "punpcklbw %%mm0, %%mm0 \n\t" - "punpcklwd %%mm0, %%mm0 \n\t" - "punpckldq %%mm0, %%mm0 \n\t" - "movq %%mm0, -8(%0) \n\t" - "movq -8(%0, %2), %%mm1 \n\t" - "punpckhbw %%mm1, %%mm1 \n\t" - "punpckhwd %%mm1, %%mm1 \n\t" - "punpckhdq %%mm1, %%mm1 \n\t" - "movq %%mm1, (%0, %2) \n\t" - "add %1, %0 \n\t" - "cmp %3, %0 \n\t" - " jb 1b \n\t" - : "+r" (ptr) - : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height) - ); + asm volatile( + "1: \n\t" + "movd (%0), %%mm0 \n\t" + "punpcklbw %%mm0, %%mm0 \n\t" + "punpcklwd %%mm0, %%mm0 \n\t" + "punpckldq %%mm0, %%mm0 \n\t" + "movq %%mm0, -8(%0) \n\t" + "movq -8(%0, %2), %%mm1 \n\t" + "punpckhbw %%mm1, %%mm1 \n\t" + "punpckhwd %%mm1, %%mm1 \n\t" + "punpckhdq %%mm1, %%mm1 \n\t" + "movq %%mm1, (%0, %2) \n\t" + "add %1, %0 \n\t" + "cmp %3, %0 \n\t" + " jb 1b \n\t" + : "+r" (ptr) + : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height) + ); } else { - asm volatile( - "1: \n\t" - "movd (%0), %%mm0 \n\t" - "punpcklbw %%mm0, %%mm0 \n\t" - "punpcklwd %%mm0, %%mm0 \n\t" - "punpckldq %%mm0, %%mm0 \n\t" - "movq %%mm0, -8(%0) \n\t" - "movq %%mm0, -16(%0) \n\t" - "movq -8(%0, %2), %%mm1 \n\t" - "punpckhbw %%mm1, %%mm1 \n\t" - "punpckhwd %%mm1, %%mm1 \n\t" - "punpckhdq %%mm1, %%mm1 \n\t" - "movq %%mm1, (%0, %2) \n\t" - "movq %%mm1, 8(%0, %2) \n\t" - "add %1, %0 \n\t" - "cmp %3, %0 \n\t" - " jb 1b \n\t" - : "+r" (ptr) - : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height) - ); + asm volatile( + "1: \n\t" + "movd (%0), %%mm0 \n\t" + "punpcklbw %%mm0, %%mm0 \n\t" + "punpcklwd %%mm0, %%mm0 \n\t" + "punpckldq %%mm0, %%mm0 \n\t" + "movq %%mm0, -8(%0) \n\t" + "movq %%mm0, -16(%0) \n\t" + "movq -8(%0, %2), %%mm1 \n\t" + "punpckhbw %%mm1, %%mm1 \n\t" + "punpckhwd %%mm1, %%mm1 \n\t" + "punpckhdq %%mm1, %%mm1 \n\t" + "movq %%mm1, (%0, %2) \n\t" + "movq %%mm1, 8(%0, %2) \n\t" + "add %1, %0 \n\t" + "cmp %3, %0 \n\t" + " jb 1b \n\t" + : "+r" (ptr) + : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height) + ); } for(i=0;i<w;i+=4) { /* top and bottom (and hopefully also the corners) */ - ptr= buf - (i + 1) * wrap - w; - asm volatile( - "1: \n\t" - "movq (%1, %0), %%mm0 \n\t" - "movq %%mm0, (%0) \n\t" - "movq %%mm0, (%0, %2) \n\t" - "movq %%mm0, (%0, %2, 2) \n\t" - "movq %%mm0, (%0, %3) \n\t" - "add $8, %0 \n\t" - "cmp %4, %0 \n\t" - " jb 1b \n\t" - : "+r" (ptr) - : "r" ((long)buf - (long)ptr - w), "r" ((long)-wrap), "r" ((long)-wrap*3), "r" (ptr+width+2*w) - ); - ptr= last_line + (i + 1) * wrap - w; - asm volatile( - "1: \n\t" - "movq (%1, %0), %%mm0 \n\t" - "movq %%mm0, (%0) \n\t" - "movq %%mm0, (%0, %2) \n\t" - "movq %%mm0, (%0, %2, 2) \n\t" - "movq %%mm0, (%0, %3) \n\t" - "add $8, %0 \n\t" - "cmp %4, %0 \n\t" - " jb 1b \n\t" - : "+r" (ptr) - : "r" ((long)last_line - (long)ptr - w), "r" ((long)wrap), "r" ((long)wrap*3), "r" (ptr+width+2*w) - ); + ptr= buf - (i + 1) * wrap - w; + asm volatile( + "1: \n\t" + "movq (%1, %0), %%mm0 \n\t" + "movq %%mm0, (%0) \n\t" + "movq %%mm0, (%0, %2) \n\t" + "movq %%mm0, (%0, %2, 2) \n\t" + "movq %%mm0, (%0, %3) \n\t" + "add $8, %0 \n\t" + "cmp %4, %0 \n\t" + " jb 1b \n\t" + : "+r" (ptr) + : "r" ((long)buf - (long)ptr - w), "r" ((long)-wrap), "r" ((long)-wrap*3), "r" (ptr+width+2*w) + ); + ptr= last_line + (i + 1) * wrap - w; + asm volatile( + "1: \n\t" + "movq (%1, %0), %%mm0 \n\t" + "movq %%mm0, (%0) \n\t" + "movq %%mm0, (%0, %2) \n\t" + "movq %%mm0, (%0, %2, 2) \n\t" + "movq %%mm0, (%0, %3) \n\t" + "add $8, %0 \n\t" + "cmp %4, %0 \n\t" + " jb 1b \n\t" + : "+r" (ptr) + : "r" ((long)last_line - (long)ptr - w), "r" ((long)wrap), "r" ((long)wrap*3), "r" (ptr+width+2*w) + ); } } @@ -572,47 +572,47 @@ s->dct_count[intra]++; asm volatile( - "pxor %%mm7, %%mm7 \n\t" - "1: \n\t" - "pxor %%mm0, %%mm0 \n\t" - "pxor %%mm1, %%mm1 \n\t" - "movq (%0), %%mm2 \n\t" - "movq 8(%0), %%mm3 \n\t" - "pcmpgtw %%mm2, %%mm0 \n\t" - "pcmpgtw %%mm3, %%mm1 \n\t" - "pxor %%mm0, %%mm2 \n\t" - "pxor %%mm1, %%mm3 \n\t" - "psubw %%mm0, %%mm2 \n\t" - "psubw %%mm1, %%mm3 \n\t" - "movq %%mm2, %%mm4 \n\t" - "movq %%mm3, %%mm5 \n\t" - "psubusw (%2), %%mm2 \n\t" - "psubusw 8(%2), %%mm3 \n\t" - "pxor %%mm0, %%mm2 \n\t" - "pxor %%mm1, %%mm3 \n\t" - "psubw %%mm0, %%mm2 \n\t" - "psubw %%mm1, %%mm3 \n\t" - "movq %%mm2, (%0) \n\t" - "movq %%mm3, 8(%0) \n\t" - "movq %%mm4, %%mm2 \n\t" - "movq %%mm5, %%mm3 \n\t" - "punpcklwd %%mm7, %%mm4 \n\t" - "punpckhwd %%mm7, %%mm2 \n\t" - "punpcklwd %%mm7, %%mm5 \n\t" - "punpckhwd %%mm7, %%mm3 \n\t" - "paddd (%1), %%mm4 \n\t" - "paddd 8(%1), %%mm2 \n\t" - "paddd 16(%1), %%mm5 \n\t" - "paddd 24(%1), %%mm3 \n\t" - "movq %%mm4, (%1) \n\t" - "movq %%mm2, 8(%1) \n\t" - "movq %%mm5, 16(%1) \n\t" - "movq %%mm3, 24(%1) \n\t" - "add $16, %0 \n\t" - "add $32, %1 \n\t" - "add $16, %2 \n\t" - "cmp %3, %0 \n\t" - " jb 1b \n\t" + "pxor %%mm7, %%mm7 \n\t" + "1: \n\t" + "pxor %%mm0, %%mm0 \n\t" + "pxor %%mm1, %%mm1 \n\t" + "movq (%0), %%mm2 \n\t" + "movq 8(%0), %%mm3 \n\t" + "pcmpgtw %%mm2, %%mm0 \n\t" + "pcmpgtw %%mm3, %%mm1 \n\t" + "pxor %%mm0, %%mm2 \n\t" + "pxor %%mm1, %%mm3 \n\t" + "psubw %%mm0, %%mm2 \n\t" + "psubw %%mm1, %%mm3 \n\t" + "movq %%mm2, %%mm4 \n\t" + "movq %%mm3, %%mm5 \n\t" + "psubusw (%2), %%mm2 \n\t" + "psubusw 8(%2), %%mm3 \n\t" + "pxor %%mm0, %%mm2 \n\t" + "pxor %%mm1, %%mm3 \n\t" + "psubw %%mm0, %%mm2 \n\t" + "psubw %%mm1, %%mm3 \n\t" + "movq %%mm2, (%0) \n\t" + "movq %%mm3, 8(%0) \n\t" + "movq %%mm4, %%mm2 \n\t" + "movq %%mm5, %%mm3 \n\t" + "punpcklwd %%mm7, %%mm4 \n\t" + "punpckhwd %%mm7, %%mm2 \n\t" + "punpcklwd %%mm7, %%mm5 \n\t" + "punpckhwd %%mm7, %%mm3 \n\t" + "paddd (%1), %%mm4 \n\t" + "paddd 8(%1), %%mm2 \n\t" + "paddd 16(%1), %%mm5 \n\t" + "paddd 24(%1), %%mm3 \n\t" + "movq %%mm4, (%1) \n\t" + "movq %%mm2, 8(%1) \n\t" + "movq %%mm5, 16(%1) \n\t" + "movq %%mm3, 24(%1) \n\t" + "add $16, %0 \n\t" + "add $32, %1 \n\t" + "add $16, %2 \n\t" + "cmp %3, %0 \n\t" + " jb 1b \n\t" : "+r" (block), "+r" (sum), "+r" (offset) : "r"(block+64) ); @@ -626,47 +626,47 @@ s->dct_count[intra]++; asm volatile( - "pxor %%xmm7, %%xmm7 \n\t" - "1: \n\t" - "pxor %%xmm0, %%xmm0 \n\t" - "pxor %%xmm1, %%xmm1 \n\t" - "movdqa (%0), %%xmm2 \n\t" - "movdqa 16(%0), %%xmm3 \n\t" - "pcmpgtw %%xmm2, %%xmm0 \n\t" - "pcmpgtw %%xmm3, %%xmm1 \n\t" - "pxor %%xmm0, %%xmm2 \n\t" - "pxor %%xmm1, %%xmm3 \n\t" - "psubw %%xmm0, %%xmm2 \n\t" - "psubw %%xmm1, %%xmm3 \n\t" - "movdqa %%xmm2, %%xmm4 \n\t" - "movdqa %%xmm3, %%xmm5 \n\t" - "psubusw (%2), %%xmm2 \n\t" - "psubusw 16(%2), %%xmm3 \n\t" - "pxor %%xmm0, %%xmm2 \n\t" - "pxor %%xmm1, %%xmm3 \n\t" - "psubw %%xmm0, %%xmm2 \n\t" - "psubw %%xmm1, %%xmm3 \n\t" - "movdqa %%xmm2, (%0) \n\t" - "movdqa %%xmm3, 16(%0) \n\t" - "movdqa %%xmm4, %%xmm6 \n\t" - "movdqa %%xmm5, %%xmm0 \n\t" - "punpcklwd %%xmm7, %%xmm4 \n\t" - "punpckhwd %%xmm7, %%xmm6 \n\t" - "punpcklwd %%xmm7, %%xmm5 \n\t" - "punpckhwd %%xmm7, %%xmm0 \n\t" - "paddd (%1), %%xmm4 \n\t" - "paddd 16(%1), %%xmm6 \n\t" - "paddd 32(%1), %%xmm5 \n\t" - "paddd 48(%1), %%xmm0 \n\t" - "movdqa %%xmm4, (%1) \n\t" - "movdqa %%xmm6, 16(%1) \n\t" - "movdqa %%xmm5, 32(%1) \n\t" - "movdqa %%xmm0, 48(%1) \n\t" - "add $32, %0 \n\t" - "add $64, %1 \n\t" - "add $32, %2 \n\t" - "cmp %3, %0 \n\t" - " jb 1b \n\t" + "pxor %%xmm7, %%xmm7 \n\t" + "1: \n\t" + "pxor %%xmm0, %%xmm0 \n\t" + "pxor %%xmm1, %%xmm1 \n\t" + "movdqa (%0), %%xmm2 \n\t" + "movdqa 16(%0), %%xmm3 \n\t" + "pcmpgtw %%xmm2, %%xmm0 \n\t" + "pcmpgtw %%xmm3, %%xmm1 \n\t" + "pxor %%xmm0, %%xmm2 \n\t" + "pxor %%xmm1, %%xmm3 \n\t" + "psubw %%xmm0, %%xmm2 \n\t" + "psubw %%xmm1, %%xmm3 \n\t" + "movdqa %%xmm2, %%xmm4 \n\t" + "movdqa %%xmm3, %%xmm5 \n\t" + "psubusw (%2), %%xmm2 \n\t" + "psubusw 16(%2), %%xmm3 \n\t" + "pxor %%xmm0, %%xmm2 \n\t" + "pxor %%xmm1, %%xmm3 \n\t" + "psubw %%xmm0, %%xmm2 \n\t" + "psubw %%xmm1, %%xmm3 \n\t" + "movdqa %%xmm2, (%0) \n\t" + "movdqa %%xmm3, 16(%0) \n\t" + "movdqa %%xmm4, %%xmm6 \n\t" + "movdqa %%xmm5, %%xmm0 \n\t" + "punpcklwd %%xmm7, %%xmm4 \n\t" + "punpckhwd %%xmm7, %%xmm6 \n\t" + "punpcklwd %%xmm7, %%xmm5 \n\t" + "punpckhwd %%xmm7, %%xmm0 \n\t" + "paddd (%1), %%xmm4 \n\t" + "paddd 16(%1), %%xmm6 \n\t" + "paddd 32(%1), %%xmm5 \n\t" + "paddd 48(%1), %%xmm0 \n\t" + "movdqa %%xmm4, (%1) \n\t" + "movdqa %%xmm6, 16(%1) \n\t" + "movdqa %%xmm5, 32(%1) \n\t" + "movdqa %%xmm0, 48(%1) \n\t" + "add $32, %0 \n\t" + "add $64, %1 \n\t" + "add $32, %2 \n\t" + "cmp %3, %0 \n\t" + " jb 1b \n\t" : "+r" (block), "+r" (sum), "+r" (offset) : "r"(block+64) ); @@ -705,10 +705,10 @@ draw_edges = draw_edges_mmx; if (mm_flags & MM_SSE2) { - s->denoise_dct= denoise_dct_sse2; - } else { - s->denoise_dct= denoise_dct_mmx; - } + s->denoise_dct= denoise_dct_sse2; + } else { + s->denoise_dct= denoise_dct_mmx; + } if(dct_algo==FF_DCT_AUTO || dct_algo==FF_DCT_MMX){ if(mm_flags & MM_SSE2){
--- a/i386/mpegvideo_mmx_template.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/mpegvideo_mmx_template.c Thu Dec 22 01:10:11 2005 +0000 @@ -21,26 +21,26 @@ #undef PMAXW #ifdef HAVE_MMX2 #define SPREADW(a) "pshufw $0, " #a ", " #a " \n\t" -#define PMAXW(a,b) "pmaxsw " #a ", " #b " \n\t" +#define PMAXW(a,b) "pmaxsw " #a ", " #b " \n\t" #define PMAX(a,b) \ - "pshufw $0x0E," #a ", " #b " \n\t"\ - PMAXW(b, a)\ - "pshufw $0x01," #a ", " #b " \n\t"\ - PMAXW(b, a) + "pshufw $0x0E," #a ", " #b " \n\t"\ + PMAXW(b, a)\ + "pshufw $0x01," #a ", " #b " \n\t"\ + PMAXW(b, a) #else #define SPREADW(a) \ - "punpcklwd " #a ", " #a " \n\t"\ - "punpcklwd " #a ", " #a " \n\t" + "punpcklwd " #a ", " #a " \n\t"\ + "punpcklwd " #a ", " #a " \n\t" #define PMAXW(a,b) \ - "psubusw " #a ", " #b " \n\t"\ - "paddw " #a ", " #b " \n\t" + "psubusw " #a ", " #b " \n\t"\ + "paddw " #a ", " #b " \n\t" #define PMAX(a,b) \ - "movq " #a ", " #b " \n\t"\ - "psrlq $32, " #a " \n\t"\ - PMAXW(b, a)\ - "movq " #a ", " #b " \n\t"\ - "psrlq $16, " #a " \n\t"\ - PMAXW(b, a) + "movq " #a ", " #b " \n\t"\ + "psrlq $32, " #a " \n\t"\ + PMAXW(b, a)\ + "movq " #a ", " #b " \n\t"\ + "psrlq $16, " #a " \n\t"\ + PMAXW(b, a) #endif @@ -71,18 +71,18 @@ if (!s->h263_aic) { #if 1 asm volatile ( - "mul %%ecx \n\t" - : "=d" (level), "=a"(dummy) - : "a" ((block[0]>>2) + q), "c" (inverse[q<<1]) + "mul %%ecx \n\t" + : "=d" (level), "=a"(dummy) + : "a" ((block[0]>>2) + q), "c" (inverse[q<<1]) ); #else asm volatile ( - "xorl %%edx, %%edx \n\t" - "divw %%cx \n\t" - "movzwl %%ax, %%eax \n\t" - : "=a" (level) - : "a" ((block[0]>>2) + q), "c" (q<<1) - : "%edx" + "xorl %%edx, %%edx \n\t" + "divw %%cx \n\t" + "movzwl %%ax, %%eax \n\t" + : "=a" (level) + : "a" ((block[0]>>2) + q), "c" (q<<1) + : "%edx" ); #endif } else @@ -103,94 +103,94 @@ if((s->out_format == FMT_H263 || s->out_format == FMT_H261) && s->mpeg_quant==0){ asm volatile( - "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1 + "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1 SPREADW(%%mm3) - "pxor %%mm7, %%mm7 \n\t" // 0 - "pxor %%mm4, %%mm4 \n\t" // 0 - "movq (%2), %%mm5 \n\t" // qmat[0] - "pxor %%mm6, %%mm6 \n\t" - "psubw (%3), %%mm6 \n\t" // -bias[0] - "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" - "1: \n\t" - "pxor %%mm1, %%mm1 \n\t" // 0 - "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] - "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00 - "pxor %%mm1, %%mm0 \n\t" - "psubw %%mm1, %%mm0 \n\t" // ABS(block[i]) - "psubusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0] - "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16 - "por %%mm0, %%mm4 \n\t" - "pxor %%mm1, %%mm0 \n\t" - "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i]) - "movq %%mm0, (%5, %%"REG_a") \n\t" - "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00 - "movq (%4, %%"REG_a"), %%mm1 \n\t" - "movq %%mm7, (%1, %%"REG_a") \n\t" // 0 - "pandn %%mm1, %%mm0 \n\t" - PMAXW(%%mm0, %%mm3) - "add $8, %%"REG_a" \n\t" - " js 1b \n\t" - PMAX(%%mm3, %%mm0) - "movd %%mm3, %%"REG_a" \n\t" - "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1 - : "+a" (last_non_zero_p1) + "pxor %%mm7, %%mm7 \n\t" // 0 + "pxor %%mm4, %%mm4 \n\t" // 0 + "movq (%2), %%mm5 \n\t" // qmat[0] + "pxor %%mm6, %%mm6 \n\t" + "psubw (%3), %%mm6 \n\t" // -bias[0] + "mov $-128, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "pxor %%mm1, %%mm1 \n\t" // 0 + "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] + "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00 + "pxor %%mm1, %%mm0 \n\t" + "psubw %%mm1, %%mm0 \n\t" // ABS(block[i]) + "psubusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0] + "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16 + "por %%mm0, %%mm4 \n\t" + "pxor %%mm1, %%mm0 \n\t" + "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i]) + "movq %%mm0, (%5, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00 + "movq (%4, %%"REG_a"), %%mm1 \n\t" + "movq %%mm7, (%1, %%"REG_a") \n\t" // 0 + "pandn %%mm1, %%mm0 \n\t" + PMAXW(%%mm0, %%mm3) + "add $8, %%"REG_a" \n\t" + " js 1b \n\t" + PMAX(%%mm3, %%mm0) + "movd %%mm3, %%"REG_a" \n\t" + "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1 + : "+a" (last_non_zero_p1) : "r" (block+64), "r" (qmat), "r" (bias), "r" (inv_zigzag_direct16+64), "r" (temp_block+64) ); // note the asm is split cuz gcc doesnt like that many operands ... asm volatile( - "movd %1, %%mm1 \n\t" // max_qcoeff - SPREADW(%%mm1) - "psubusw %%mm1, %%mm4 \n\t" - "packuswb %%mm4, %%mm4 \n\t" - "movd %%mm4, %0 \n\t" // *overflow + "movd %1, %%mm1 \n\t" // max_qcoeff + SPREADW(%%mm1) + "psubusw %%mm1, %%mm4 \n\t" + "packuswb %%mm4, %%mm4 \n\t" + "movd %%mm4, %0 \n\t" // *overflow : "=g" (*overflow) : "g" (s->max_qcoeff) ); }else{ // FMT_H263 asm volatile( - "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1 + "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1 SPREADW(%%mm3) - "pxor %%mm7, %%mm7 \n\t" // 0 - "pxor %%mm4, %%mm4 \n\t" // 0 - "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" - "1: \n\t" - "pxor %%mm1, %%mm1 \n\t" // 0 - "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] - "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00 - "pxor %%mm1, %%mm0 \n\t" - "psubw %%mm1, %%mm0 \n\t" // ABS(block[i]) - "movq (%3, %%"REG_a"), %%mm6 \n\t" // bias[0] - "paddusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0] - "movq (%2, %%"REG_a"), %%mm5 \n\t" // qmat[i] - "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] + bias[0]*qmat[0])>>16 - "por %%mm0, %%mm4 \n\t" - "pxor %%mm1, %%mm0 \n\t" - "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i]) - "movq %%mm0, (%5, %%"REG_a") \n\t" - "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00 - "movq (%4, %%"REG_a"), %%mm1 \n\t" - "movq %%mm7, (%1, %%"REG_a") \n\t" // 0 - "pandn %%mm1, %%mm0 \n\t" - PMAXW(%%mm0, %%mm3) - "add $8, %%"REG_a" \n\t" - " js 1b \n\t" - PMAX(%%mm3, %%mm0) - "movd %%mm3, %%"REG_a" \n\t" - "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1 - : "+a" (last_non_zero_p1) + "pxor %%mm7, %%mm7 \n\t" // 0 + "pxor %%mm4, %%mm4 \n\t" // 0 + "mov $-128, %%"REG_a" \n\t" + ".balign 16 \n\t" + "1: \n\t" + "pxor %%mm1, %%mm1 \n\t" // 0 + "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] + "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00 + "pxor %%mm1, %%mm0 \n\t" + "psubw %%mm1, %%mm0 \n\t" // ABS(block[i]) + "movq (%3, %%"REG_a"), %%mm6 \n\t" // bias[0] + "paddusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0] + "movq (%2, %%"REG_a"), %%mm5 \n\t" // qmat[i] + "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] + bias[0]*qmat[0])>>16 + "por %%mm0, %%mm4 \n\t" + "pxor %%mm1, %%mm0 \n\t" + "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i]) + "movq %%mm0, (%5, %%"REG_a") \n\t" + "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00 + "movq (%4, %%"REG_a"), %%mm1 \n\t" + "movq %%mm7, (%1, %%"REG_a") \n\t" // 0 + "pandn %%mm1, %%mm0 \n\t" + PMAXW(%%mm0, %%mm3) + "add $8, %%"REG_a" \n\t" + " js 1b \n\t" + PMAX(%%mm3, %%mm0) + "movd %%mm3, %%"REG_a" \n\t" + "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1 + : "+a" (last_non_zero_p1) : "r" (block+64), "r" (qmat+64), "r" (bias+64), "r" (inv_zigzag_direct16+64), "r" (temp_block+64) ); // note the asm is split cuz gcc doesnt like that many operands ... asm volatile( - "movd %1, %%mm1 \n\t" // max_qcoeff - SPREADW(%%mm1) - "psubusw %%mm1, %%mm4 \n\t" - "packuswb %%mm4, %%mm4 \n\t" - "movd %%mm4, %0 \n\t" // *overflow + "movd %1, %%mm1 \n\t" // max_qcoeff + SPREADW(%%mm1) + "psubusw %%mm1, %%mm4 \n\t" + "packuswb %%mm4, %%mm4 \n\t" + "movd %%mm4, %0 \n\t" // *overflow : "=g" (*overflow) : "g" (s->max_qcoeff) );
--- a/i386/simple_idct_mmx.c Wed Dec 21 17:50:40 2005 +0000 +++ b/i386/simple_idct_mmx.c Thu Dec 22 01:10:11 2005 +0000 @@ -40,8 +40,8 @@ #define C4 16383 //cos(i*M_PI/16)*sqrt(2)*(1<<14) - 0.5 #endif #define C5 12873 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 -#define C6 8867 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 -#define C7 4520 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define C6 8867 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define C7 4520 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 #define ROW_SHIFT 11 #define COL_SHIFT 20 // 6 @@ -50,13 +50,13 @@ static const uint64_t attribute_used __attribute__((aligned(8))) d40000= 0x0000000000040000ULL; static const int16_t __attribute__((aligned(8))) coeffs[]= { - 1<<(ROW_SHIFT-1), 0, 1<<(ROW_SHIFT-1), 0, -// 1<<(COL_SHIFT-1), 0, 1<<(COL_SHIFT-1), 0, -// 0, 1<<(COL_SHIFT-1-16), 0, 1<<(COL_SHIFT-1-16), - 1<<(ROW_SHIFT-1), 1, 1<<(ROW_SHIFT-1), 0, - // the 1 = ((1<<(COL_SHIFT-1))/C4)<<ROW_SHIFT :) -// 0, 0, 0, 0, -// 0, 0, 0, 0, + 1<<(ROW_SHIFT-1), 0, 1<<(ROW_SHIFT-1), 0, +// 1<<(COL_SHIFT-1), 0, 1<<(COL_SHIFT-1), 0, +// 0, 1<<(COL_SHIFT-1-16), 0, 1<<(COL_SHIFT-1-16), + 1<<(ROW_SHIFT-1), 1, 1<<(ROW_SHIFT-1), 0, + // the 1 = ((1<<(COL_SHIFT-1))/C4)<<ROW_SHIFT :) +// 0, 0, 0, 0, +// 0, 0, 0, 0, C4, C4, C4, C4, C4, -C4, C4, -C4, @@ -79,8 +79,8 @@ #if 0 static void unused_var_killer(){ - int a= wm1010 + d40000; - temp[0]=a; + int a= wm1010 + d40000; + temp[0]=a; } static void inline idctCol (int16_t * col, int16_t *input) @@ -93,21 +93,21 @@ #undef C5 #undef C6 #undef C7 - int a0, a1, a2, a3, b0, b1, b2, b3; - const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + int a0, a1, a2, a3, b0, b1, b2, b3; + const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 /* - if( !(col[8*1] | col[8*2] |col[8*3] |col[8*4] |col[8*5] |col[8*6] | col[8*7])) { - col[8*0] = col[8*1] = col[8*2] = col[8*3] = col[8*4] = - col[8*5] = col[8*6] = col[8*7] = col[8*0]<<3; - return; - }*/ + if( !(col[8*1] | col[8*2] |col[8*3] |col[8*4] |col[8*5] |col[8*6] | col[8*7])) { + col[8*0] = col[8*1] = col[8*2] = col[8*3] = col[8*4] = + col[8*5] = col[8*6] = col[8*7] = col[8*0]<<3; + return; + }*/ col[8*0] = input[8*0 + 0]; col[8*1] = input[8*2 + 0]; @@ -118,39 +118,39 @@ col[8*6] = input[8*4 + 1]; col[8*7] = input[8*6 + 1]; - a0 = C4*col[8*0] + C2*col[8*2] + C4*col[8*4] + C6*col[8*6] + (1<<(COL_SHIFT-1)); - a1 = C4*col[8*0] + C6*col[8*2] - C4*col[8*4] - C2*col[8*6] + (1<<(COL_SHIFT-1)); - a2 = C4*col[8*0] - C6*col[8*2] - C4*col[8*4] + C2*col[8*6] + (1<<(COL_SHIFT-1)); - a3 = C4*col[8*0] - C2*col[8*2] + C4*col[8*4] - C6*col[8*6] + (1<<(COL_SHIFT-1)); + a0 = C4*col[8*0] + C2*col[8*2] + C4*col[8*4] + C6*col[8*6] + (1<<(COL_SHIFT-1)); + a1 = C4*col[8*0] + C6*col[8*2] - C4*col[8*4] - C2*col[8*6] + (1<<(COL_SHIFT-1)); + a2 = C4*col[8*0] - C6*col[8*2] - C4*col[8*4] + C2*col[8*6] + (1<<(COL_SHIFT-1)); + a3 = C4*col[8*0] - C2*col[8*2] + C4*col[8*4] - C6*col[8*6] + (1<<(COL_SHIFT-1)); - b0 = C1*col[8*1] + C3*col[8*3] + C5*col[8*5] + C7*col[8*7]; - b1 = C3*col[8*1] - C7*col[8*3] - C1*col[8*5] - C5*col[8*7]; - b2 = C5*col[8*1] - C1*col[8*3] + C7*col[8*5] + C3*col[8*7]; - b3 = C7*col[8*1] - C5*col[8*3] + C3*col[8*5] - C1*col[8*7]; + b0 = C1*col[8*1] + C3*col[8*3] + C5*col[8*5] + C7*col[8*7]; + b1 = C3*col[8*1] - C7*col[8*3] - C1*col[8*5] - C5*col[8*7]; + b2 = C5*col[8*1] - C1*col[8*3] + C7*col[8*5] + C3*col[8*7]; + b3 = C7*col[8*1] - C5*col[8*3] + C3*col[8*5] - C1*col[8*7]; - col[8*0] = (a0 + b0) >> COL_SHIFT; - col[8*1] = (a1 + b1) >> COL_SHIFT; - col[8*2] = (a2 + b2) >> COL_SHIFT; - col[8*3] = (a3 + b3) >> COL_SHIFT; - col[8*4] = (a3 - b3) >> COL_SHIFT; - col[8*5] = (a2 - b2) >> COL_SHIFT; - col[8*6] = (a1 - b1) >> COL_SHIFT; - col[8*7] = (a0 - b0) >> COL_SHIFT; + col[8*0] = (a0 + b0) >> COL_SHIFT; + col[8*1] = (a1 + b1) >> COL_SHIFT; + col[8*2] = (a2 + b2) >> COL_SHIFT; + col[8*3] = (a3 + b3) >> COL_SHIFT; + col[8*4] = (a3 - b3) >> COL_SHIFT; + col[8*5] = (a2 - b2) >> COL_SHIFT; + col[8*6] = (a1 - b1) >> COL_SHIFT; + col[8*7] = (a0 - b0) >> COL_SHIFT; } static void inline idctRow (int16_t * output, int16_t * input) { - int16_t row[8]; + int16_t row[8]; - int a0, a1, a2, a3, b0, b1, b2, b3; - const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 - const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + int a0, a1, a2, a3, b0, b1, b2, b3; + const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 + const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 row[0] = input[0]; row[2] = input[1]; @@ -161,290 +161,290 @@ row[5] = input[12]; row[7] = input[13]; - if( !(row[1] | row[2] |row[3] |row[4] |row[5] |row[6] | row[7]) ) { - row[0] = row[1] = row[2] = row[3] = row[4] = - row[5] = row[6] = row[7] = row[0]<<3; - output[0] = row[0]; - output[2] = row[1]; - output[4] = row[2]; - output[6] = row[3]; - output[8] = row[4]; - output[10] = row[5]; - output[12] = row[6]; - output[14] = row[7]; - return; - } + if( !(row[1] | row[2] |row[3] |row[4] |row[5] |row[6] | row[7]) ) { + row[0] = row[1] = row[2] = row[3] = row[4] = + row[5] = row[6] = row[7] = row[0]<<3; + output[0] = row[0]; + output[2] = row[1]; + output[4] = row[2]; + output[6] = row[3]; + output[8] = row[4]; + output[10] = row[5]; + output[12] = row[6]; + output[14] = row[7]; + return; + } - a0 = C4*row[0] + C2*row[2] + C4*row[4] + C6*row[6] + (1<<(ROW_SHIFT-1)); - a1 = C4*row[0] + C6*row[2] - C4*row[4] - C2*row[6] + (1<<(ROW_SHIFT-1)); - a2 = C4*row[0] - C6*row[2] - C4*row[4] + C2*row[6] + (1<<(ROW_SHIFT-1)); - a3 = C4*row[0] - C2*row[2] + C4*row[4] - C6*row[6] + (1<<(ROW_SHIFT-1)); + a0 = C4*row[0] + C2*row[2] + C4*row[4] + C6*row[6] + (1<<(ROW_SHIFT-1)); + a1 = C4*row[0] + C6*row[2] - C4*row[4] - C2*row[6] + (1<<(ROW_SHIFT-1)); + a2 = C4*row[0] - C6*row[2] - C4*row[4] + C2*row[6] + (1<<(ROW_SHIFT-1)); + a3 = C4*row[0] - C2*row[2] + C4*row[4] - C6*row[6] + (1<<(ROW_SHIFT-1)); - b0 = C1*row[1] + C3*row[3] + C5*row[5] + C7*row[7]; - b1 = C3*row[1] - C7*row[3] - C1*row[5] - C5*row[7]; - b2 = C5*row[1] - C1*row[3] + C7*row[5] + C3*row[7]; - b3 = C7*row[1] - C5*row[3] + C3*row[5] - C1*row[7]; + b0 = C1*row[1] + C3*row[3] + C5*row[5] + C7*row[7]; + b1 = C3*row[1] - C7*row[3] - C1*row[5] - C5*row[7]; + b2 = C5*row[1] - C1*row[3] + C7*row[5] + C3*row[7]; + b3 = C7*row[1] - C5*row[3] + C3*row[5] - C1*row[7]; - row[0] = (a0 + b0) >> ROW_SHIFT; - row[1] = (a1 + b1) >> ROW_SHIFT; - row[2] = (a2 + b2) >> ROW_SHIFT; - row[3] = (a3 + b3) >> ROW_SHIFT; - row[4] = (a3 - b3) >> ROW_SHIFT; - row[5] = (a2 - b2) >> ROW_SHIFT; - row[6] = (a1 - b1) >> ROW_SHIFT; - row[7] = (a0 - b0) >> ROW_SHIFT; + row[0] = (a0 + b0) >> ROW_SHIFT; + row[1] = (a1 + b1) >> ROW_SHIFT; + row[2] = (a2 + b2) >> ROW_SHIFT; + row[3] = (a3 + b3) >> ROW_SHIFT; + row[4] = (a3 - b3) >> ROW_SHIFT; + row[5] = (a2 - b2) >> ROW_SHIFT; + row[6] = (a1 - b1) >> ROW_SHIFT; + row[7] = (a0 - b0) >> ROW_SHIFT; - output[0] = row[0]; - output[2] = row[1]; - output[4] = row[2]; - output[6] = row[3]; - output[8] = row[4]; - output[10] = row[5]; - output[12] = row[6]; - output[14] = row[7]; + output[0] = row[0]; + output[2] = row[1]; + output[4] = row[2]; + output[6] = row[3]; + output[8] = row[4]; + output[10] = row[5]; + output[12] = row[6]; + output[14] = row[7]; } #endif static inline void idct(int16_t *block) { - int64_t __attribute__((aligned(8))) align_tmp[16]; - int16_t * const temp= (int16_t*)align_tmp; + int64_t __attribute__((aligned(8))) align_tmp[16]; + int16_t * const temp= (int16_t*)align_tmp; - asm volatile( + asm volatile( #if 0 //Alternative, simpler variant #define ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm0, %%mm0 \n\t" \ - "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ - "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ - "movq %%mm7, " #dst " \n\t"\ - "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "movq %%mm2, 24+" #dst " \n\t"\ - "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ - "movq %%mm2, 8+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ - "movq %%mm4, 16+" #dst " \n\t"\ + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm0, %%mm0 \n\t" \ + "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ + "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ + "movq %%mm7, " #dst " \n\t"\ + "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "movq %%mm2, 24+" #dst " \n\t"\ + "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ + "movq %%mm2, 8+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ + "movq %%mm4, 16+" #dst " \n\t"\ #define COL_IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - #rounder ", %%mm0 \n\t"\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ - "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ - "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm7, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm2, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm2, 32+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm4, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t"\ + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + #rounder ", %%mm0 \n\t"\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ + "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ + "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm7, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm2, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm2, 32+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm4, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t"\ #define DC_COND_ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq "MANGLE(wm1010)", %%mm4 \n\t"\ - "pand %%mm0, %%mm4 \n\t"\ - "por %%mm1, %%mm4 \n\t"\ - "por %%mm2, %%mm4 \n\t"\ - "por %%mm3, %%mm4 \n\t"\ - "packssdw %%mm4,%%mm4 \n\t"\ - "movd %%mm4, %%eax \n\t"\ - "orl %%eax, %%eax \n\t"\ - "jz 1f \n\t"\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm0, %%mm0 \n\t" \ - "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ - "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ - "movq %%mm7, " #dst " \n\t"\ - "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "movq %%mm2, 24+" #dst " \n\t"\ - "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ - "movq %%mm2, 8+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ - "movq %%mm4, 16+" #dst " \n\t"\ - "jmp 2f \n\t"\ - "1: \n\t"\ - "pslld $16, %%mm0 \n\t"\ - "#paddd "MANGLE(d40000)", %%mm0 \n\t"\ - "psrad $13, %%mm0 \n\t"\ - "packssdw %%mm0, %%mm0 \n\t"\ - "movq %%mm0, " #dst " \n\t"\ - "movq %%mm0, 8+" #dst " \n\t"\ - "movq %%mm0, 16+" #dst " \n\t"\ - "movq %%mm0, 24+" #dst " \n\t"\ - "2: \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq "MANGLE(wm1010)", %%mm4 \n\t"\ + "pand %%mm0, %%mm4 \n\t"\ + "por %%mm1, %%mm4 \n\t"\ + "por %%mm2, %%mm4 \n\t"\ + "por %%mm3, %%mm4 \n\t"\ + "packssdw %%mm4,%%mm4 \n\t"\ + "movd %%mm4, %%eax \n\t"\ + "orl %%eax, %%eax \n\t"\ + "jz 1f \n\t"\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm0, %%mm0 \n\t" \ + "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ + "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ + "movq %%mm7, " #dst " \n\t"\ + "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "movq %%mm2, 24+" #dst " \n\t"\ + "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ + "movq %%mm2, 8+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ + "movq %%mm4, 16+" #dst " \n\t"\ + "jmp 2f \n\t"\ + "1: \n\t"\ + "pslld $16, %%mm0 \n\t"\ + "#paddd "MANGLE(d40000)", %%mm0 \n\t"\ + "psrad $13, %%mm0 \n\t"\ + "packssdw %%mm0, %%mm0 \n\t"\ + "movq %%mm0, " #dst " \n\t"\ + "movq %%mm0, 8+" #dst " \n\t"\ + "movq %%mm0, 16+" #dst " \n\t"\ + "movq %%mm0, 24+" #dst " \n\t"\ + "2: \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -467,236 +467,236 @@ #else #define DC_COND_IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq "MANGLE(wm1010)", %%mm4 \n\t"\ - "pand %%mm0, %%mm4 \n\t"\ - "por %%mm1, %%mm4 \n\t"\ - "por %%mm2, %%mm4 \n\t"\ - "por %%mm3, %%mm4 \n\t"\ - "packssdw %%mm4,%%mm4 \n\t"\ - "movd %%mm4, %%eax \n\t"\ - "orl %%eax, %%eax \n\t"\ - "jz 1f \n\t"\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm0, %%mm0 \n\t" \ - "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ - "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ - "movq %%mm7, " #dst " \n\t"\ - "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "movq %%mm2, 24+" #dst " \n\t"\ - "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ - "movq %%mm2, 8+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ - "movq %%mm4, 16+" #dst " \n\t"\ - "jmp 2f \n\t"\ - "1: \n\t"\ - "pslld $16, %%mm0 \n\t"\ - "paddd "MANGLE(d40000)", %%mm0 \n\t"\ - "psrad $13, %%mm0 \n\t"\ - "packssdw %%mm0, %%mm0 \n\t"\ - "movq %%mm0, " #dst " \n\t"\ - "movq %%mm0, 8+" #dst " \n\t"\ - "movq %%mm0, 16+" #dst " \n\t"\ - "movq %%mm0, 24+" #dst " \n\t"\ - "2: \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq "MANGLE(wm1010)", %%mm4 \n\t"\ + "pand %%mm0, %%mm4 \n\t"\ + "por %%mm1, %%mm4 \n\t"\ + "por %%mm2, %%mm4 \n\t"\ + "por %%mm3, %%mm4 \n\t"\ + "packssdw %%mm4,%%mm4 \n\t"\ + "movd %%mm4, %%eax \n\t"\ + "orl %%eax, %%eax \n\t"\ + "jz 1f \n\t"\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm0, %%mm0 \n\t" \ + "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ + "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ + "movq %%mm7, " #dst " \n\t"\ + "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "movq %%mm2, 24+" #dst " \n\t"\ + "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ + "movq %%mm2, 8+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ + "movq %%mm4, 16+" #dst " \n\t"\ + "jmp 2f \n\t"\ + "1: \n\t"\ + "pslld $16, %%mm0 \n\t"\ + "paddd "MANGLE(d40000)", %%mm0 \n\t"\ + "psrad $13, %%mm0 \n\t"\ + "packssdw %%mm0, %%mm0 \n\t"\ + "movq %%mm0, " #dst " \n\t"\ + "movq %%mm0, 8+" #dst " \n\t"\ + "movq %%mm0, 16+" #dst " \n\t"\ + "movq %%mm0, 24+" #dst " \n\t"\ + "2: \n\t" #define Z_COND_IDCT(src0, src4, src1, src5, dst, rounder, shift, bt) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq %%mm0, %%mm4 \n\t"\ - "por %%mm1, %%mm4 \n\t"\ - "por %%mm2, %%mm4 \n\t"\ - "por %%mm3, %%mm4 \n\t"\ - "packssdw %%mm4,%%mm4 \n\t"\ - "movd %%mm4, %%eax \n\t"\ - "orl %%eax, %%eax \n\t"\ - "jz " #bt " \n\t"\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm0, %%mm0 \n\t" \ - "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ - "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ - "movq %%mm7, " #dst " \n\t"\ - "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "movq %%mm2, 24+" #dst " \n\t"\ - "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ - "movq %%mm2, 8+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ - "movq %%mm4, 16+" #dst " \n\t"\ + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq %%mm0, %%mm4 \n\t"\ + "por %%mm1, %%mm4 \n\t"\ + "por %%mm2, %%mm4 \n\t"\ + "por %%mm3, %%mm4 \n\t"\ + "packssdw %%mm4,%%mm4 \n\t"\ + "movd %%mm4, %%eax \n\t"\ + "orl %%eax, %%eax \n\t"\ + "jz " #bt " \n\t"\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm0, %%mm0 \n\t" \ + "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ + "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ + "movq %%mm7, " #dst " \n\t"\ + "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "movq %%mm2, 24+" #dst " \n\t"\ + "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ + "movq %%mm2, 8+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ + "movq %%mm4, 16+" #dst " \n\t"\ #define ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm0, %%mm0 \n\t" \ - "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ - "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ - "movq %%mm7, " #dst " \n\t"\ - "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "movq %%mm2, 24+" #dst " \n\t"\ - "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ - "movq %%mm2, 8+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ - "movq %%mm4, 16+" #dst " \n\t"\ + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm0, %%mm0 \n\t" \ + "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\ + "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\ + "movq %%mm7, " #dst " \n\t"\ + "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "movq %%mm2, 24+" #dst " \n\t"\ + "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\ + "movq %%mm2, 8+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\ + "movq %%mm4, 16+" #dst " \n\t"\ //IDCT( src0, src4, src1, src5, dst, rounder, shift) DC_COND_IDCT( 0(%0), 8(%0), 16(%0), 24(%0), 0(%1),paddd 8(%2), 11) @@ -706,80 +706,80 @@ #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - #rounder ", %%mm0 \n\t"\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ - "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ - "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm7, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm2, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm2, 32+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm4, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + #rounder ", %%mm0 \n\t"\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ + "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ + "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm7, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm2, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm2, 32+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm4, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -787,144 +787,144 @@ IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "4: \n\t" + "#.balign 16 \n\t"\ + "4: \n\t" Z_COND_IDCT( 64(%0), 72(%0), 80(%0), 88(%0), 64(%1),paddd (%2), 11, 6f) Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 5f) #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - #rounder ", %%mm0 \n\t"\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ - "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ - "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm1, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm2, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm1 \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm2, 32+" #dst " \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm1, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + #rounder ", %%mm0 \n\t"\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ + "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ + "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm1, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm2, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm1 \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm2, 32+" #dst " \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm1, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20) IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "6: \n\t" + "#.balign 16 \n\t"\ + "6: \n\t" Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 7f) #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - #rounder ", %%mm0 \n\t"\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm1, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm2, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm1 \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm2, 32+" #dst " \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm1, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + #rounder ", %%mm0 \n\t"\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm1, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm2, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm1 \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm2, 32+" #dst " \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm1, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -932,144 +932,144 @@ IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "2: \n\t" + "#.balign 16 \n\t"\ + "2: \n\t" Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 3f) #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - #rounder ", %%mm0 \n\t"\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ - "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ - "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ - "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm7, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm2, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ - "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ - "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ - "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ - "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ - "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ - "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm2, 32+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm4, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + #rounder ", %%mm0 \n\t"\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\ + "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\ + "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\ + "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm7, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm2, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\ + "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\ + "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\ + "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\ + "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\ + "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\ + "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm2, 32+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm4, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20) IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "3: \n\t" + "#.balign 16 \n\t"\ + "3: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - #rounder ", %%mm0 \n\t"\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 64(%2), %%mm3 \n\t"\ - "pmaddwd %%mm2, %%mm3 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm1 \n\t" /* A1 a1 */\ - "paddd %%mm3, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm3, %%mm1 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm1 \n\t"\ - "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm7, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm1, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "movq %%mm5, %%mm1 \n\t" /* A2 a2 */\ - "paddd %%mm4, %%mm1 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm1 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm1, %%mm1 \n\t" /* A2+B2 a2+b2 */\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm1, 32+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "movd %%mm4, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + #rounder ", %%mm0 \n\t"\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 64(%2), %%mm3 \n\t"\ + "pmaddwd %%mm2, %%mm3 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm1 \n\t" /* A1 a1 */\ + "paddd %%mm3, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm3, %%mm1 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm1 \n\t"\ + "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm7, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm1, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "movq %%mm5, %%mm1 \n\t" /* A2 a2 */\ + "paddd %%mm4, %%mm1 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm1 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm1, %%mm1 \n\t" /* A2+B2 a2+b2 */\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm1, 32+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "movd %%mm4, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -1077,67 +1077,67 @@ IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "5: \n\t" + "#.balign 16 \n\t"\ + "5: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - #rounder ", %%mm0 \n\t"\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ - "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ - "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\ - "movq 8+" #src4 ", %%mm3 \n\t" /* R6 R2 r6 r2 */\ - "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm3, %%mm7 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "pmaddwd 40(%2), %%mm3 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm1 \n\t"\ - "paddd %%mm1, %%mm7 \n\t" /* A0 a0 */\ - "paddd %%mm1, %%mm1 \n\t" /* 2C0 2c0 */\ - #rounder ", %%mm2 \n\t"\ - "psubd %%mm7, %%mm1 \n\t" /* A3 a3 */\ - "paddd %%mm2, %%mm3 \n\t" /* A1 a1 */\ - "paddd %%mm2, %%mm2 \n\t" /* 2C1 2c1 */\ - "psubd %%mm3, %%mm2 \n\t" /* A2 a2 */\ - "psrad $" #shift ", %%mm4 \n\t"\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm3 \n\t"\ - "packssdw %%mm7, %%mm4 \n\t" /* A0 a0 */\ - "movq %%mm4, " #dst " \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "packssdw %%mm3, %%mm0 \n\t" /* A1 a1 */\ - "movq %%mm0, 16+" #dst " \n\t"\ - "movq %%mm0, 96+" #dst " \n\t"\ - "movq %%mm4, 112+" #dst " \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "psrad $" #shift ", %%mm6 \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm2, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movq %%mm5, 32+" #dst " \n\t"\ - "psrad $" #shift ", %%mm1 \n\t"\ - "packssdw %%mm1, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movq %%mm6, 48+" #dst " \n\t"\ - "movq %%mm6, 64+" #dst " \n\t"\ - "movq %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + #rounder ", %%mm0 \n\t"\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ + "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ + "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\ + "movq 8+" #src4 ", %%mm3 \n\t" /* R6 R2 r6 r2 */\ + "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm3, %%mm7 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "pmaddwd 40(%2), %%mm3 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm1 \n\t"\ + "paddd %%mm1, %%mm7 \n\t" /* A0 a0 */\ + "paddd %%mm1, %%mm1 \n\t" /* 2C0 2c0 */\ + #rounder ", %%mm2 \n\t"\ + "psubd %%mm7, %%mm1 \n\t" /* A3 a3 */\ + "paddd %%mm2, %%mm3 \n\t" /* A1 a1 */\ + "paddd %%mm2, %%mm2 \n\t" /* 2C1 2c1 */\ + "psubd %%mm3, %%mm2 \n\t" /* A2 a2 */\ + "psrad $" #shift ", %%mm4 \n\t"\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm3 \n\t"\ + "packssdw %%mm7, %%mm4 \n\t" /* A0 a0 */\ + "movq %%mm4, " #dst " \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "packssdw %%mm3, %%mm0 \n\t" /* A1 a1 */\ + "movq %%mm0, 16+" #dst " \n\t"\ + "movq %%mm0, 96+" #dst " \n\t"\ + "movq %%mm4, 112+" #dst " \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "psrad $" #shift ", %%mm6 \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm2, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movq %%mm5, 32+" #dst " \n\t"\ + "psrad $" #shift ", %%mm1 \n\t"\ + "packssdw %%mm1, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movq %%mm6, 48+" #dst " \n\t"\ + "movq %%mm6, 64+" #dst " \n\t"\ + "movq %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -1145,75 +1145,75 @@ //IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) //IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t"\ - "1: \n\t" + "#.balign 16 \n\t"\ + "1: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ - "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ - "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ - "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ - "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ - #rounder ", %%mm4 \n\t"\ - "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ - #rounder ", %%mm0 \n\t"\ - "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ - "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ - "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ - "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ - "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ - "movq 64(%2), %%mm1 \n\t"\ - "pmaddwd %%mm2, %%mm1 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ - "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ - "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "psrad $" #shift ", %%mm7 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "movq %%mm0, %%mm3 \n\t" /* A1 a1 */\ - "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "psubd %%mm1, %%mm3 \n\t" /* A1-B1 a1-b1 */\ - "psrad $" #shift ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm3 \n\t"\ - "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ - "movd %%mm7, " #dst " \n\t"\ - "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ - "movd %%mm0, 16+" #dst " \n\t"\ - "packssdw %%mm3, %%mm3 \n\t" /* A1-B1 a1-b1 */\ - "movd %%mm3, 96+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ - "movd %%mm4, 112+" #dst " \n\t"\ - "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ - "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ - "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ - "movq %%mm5, %%mm3 \n\t" /* A2 a2 */\ - "paddd %%mm4, %%mm3 \n\t" /* A2+B2 a2+b2 */\ - "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ - "psrad $" #shift ", %%mm3 \n\t"\ - "psrad $" #shift ", %%mm5 \n\t"\ - "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ - "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\ - "psrad $" #shift ", %%mm6 \n\t"\ - "packssdw %%mm3, %%mm3 \n\t" /* A2+B2 a2+b2 */\ - "movd %%mm3, 32+" #dst " \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ - "movd %%mm6, 48+" #dst " \n\t"\ - "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ - "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ - "movd %%mm4, 64+" #dst " \n\t"\ - "movd %%mm5, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\ + "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\ + "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\ + "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\ + "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\ + #rounder ", %%mm4 \n\t"\ + "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\ + #rounder ", %%mm0 \n\t"\ + "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\ + "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\ + "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\ + "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\ + "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\ + "movq 64(%2), %%mm1 \n\t"\ + "pmaddwd %%mm2, %%mm1 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\ + "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\ + "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "psrad $" #shift ", %%mm7 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "movq %%mm0, %%mm3 \n\t" /* A1 a1 */\ + "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "psubd %%mm1, %%mm3 \n\t" /* A1-B1 a1-b1 */\ + "psrad $" #shift ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm3 \n\t"\ + "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\ + "movd %%mm7, " #dst " \n\t"\ + "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\ + "movd %%mm0, 16+" #dst " \n\t"\ + "packssdw %%mm3, %%mm3 \n\t" /* A1-B1 a1-b1 */\ + "movd %%mm3, 96+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\ + "movd %%mm4, 112+" #dst " \n\t"\ + "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\ + "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\ + "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\ + "movq %%mm5, %%mm3 \n\t" /* A2 a2 */\ + "paddd %%mm4, %%mm3 \n\t" /* A2+B2 a2+b2 */\ + "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\ + "psrad $" #shift ", %%mm3 \n\t"\ + "psrad $" #shift ", %%mm5 \n\t"\ + "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\ + "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\ + "psrad $" #shift ", %%mm6 \n\t"\ + "packssdw %%mm3, %%mm3 \n\t" /* A2+B2 a2+b2 */\ + "movd %%mm3, 32+" #dst " \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\ + "movd %%mm6, 48+" #dst " \n\t"\ + "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\ + "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\ + "movd %%mm4, 64+" #dst " \n\t"\ + "movd %%mm5, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) @@ -1221,42 +1221,42 @@ IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20) IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20) IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20) - "jmp 9f \n\t" + "jmp 9f \n\t" - "#.balign 16 \n\t" - "7: \n\t" + "#.balign 16 \n\t" + "7: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, rounder, shift) \ - "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ - "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - #rounder ", %%mm4 \n\t"\ - #rounder ", %%mm0 \n\t"\ - "psrad $" #shift ", %%mm4 \n\t"\ - "psrad $" #shift ", %%mm0 \n\t"\ - "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\ - "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\ - "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ - "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\ - "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ - "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\ - #rounder ", %%mm1 \n\t"\ - #rounder ", %%mm2 \n\t"\ - "psrad $" #shift ", %%mm1 \n\t"\ - "packssdw %%mm1, %%mm4 \n\t" /* A0 a0 */\ - "movq %%mm4, " #dst " \n\t"\ - "psrad $" #shift ", %%mm2 \n\t"\ - "packssdw %%mm2, %%mm0 \n\t" /* A1 a1 */\ - "movq %%mm0, 16+" #dst " \n\t"\ - "movq %%mm0, 96+" #dst " \n\t"\ - "movq %%mm4, 112+" #dst " \n\t"\ - "movq %%mm0, 32+" #dst " \n\t"\ - "movq %%mm4, 48+" #dst " \n\t"\ - "movq %%mm4, 64+" #dst " \n\t"\ - "movq %%mm0, 80+" #dst " \n\t" + "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\ + "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + #rounder ", %%mm4 \n\t"\ + #rounder ", %%mm0 \n\t"\ + "psrad $" #shift ", %%mm4 \n\t"\ + "psrad $" #shift ", %%mm0 \n\t"\ + "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\ + "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\ + "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\ + "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\ + "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\ + "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\ + #rounder ", %%mm1 \n\t"\ + #rounder ", %%mm2 \n\t"\ + "psrad $" #shift ", %%mm1 \n\t"\ + "packssdw %%mm1, %%mm4 \n\t" /* A0 a0 */\ + "movq %%mm4, " #dst " \n\t"\ + "psrad $" #shift ", %%mm2 \n\t"\ + "packssdw %%mm2, %%mm0 \n\t" /* A1 a1 */\ + "movq %%mm0, 16+" #dst " \n\t"\ + "movq %%mm0, 96+" #dst " \n\t"\ + "movq %%mm4, 112+" #dst " \n\t"\ + "movq %%mm0, 32+" #dst " \n\t"\ + "movq %%mm4, 48+" #dst " \n\t"\ + "movq %%mm4, 64+" #dst " \n\t"\ + "movq %%mm0, 80+" #dst " \n\t" //IDCT( src0, src4, src1, src5, dst, rounder, shift) IDCT( 0(%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20) @@ -1290,9 +1290,9 @@ */ "9: \n\t" - :: "r" (block), "r" (temp), "r" (coeffs) - : "%eax" - ); + :: "r" (block), "r" (temp), "r" (coeffs) + : "%eax" + ); } void ff_simple_idct_mmx(int16_t *block)
--- a/imgconvert.c Wed Dec 21 17:50:40 2005 +0000 +++ b/imgconvert.c Thu Dec 22 01:10:11 2005 +0000 @@ -257,13 +257,13 @@ for (i=0; i < PIX_FMT_NB; i++) if (!strcmp(pix_fmt_info[i].name, name)) - break; + break; return i; } /* Picture field are filled with 'ptr' addresses. Also return size */ int avpicture_fill(AVPicture *picture, uint8_t *ptr, - int pix_fmt, int width, int height) + int pix_fmt, int width, int height) { int size, w2, h2, size2; PixFmtInfo *pinfo; @@ -373,36 +373,36 @@ pix_fmt == PIX_FMT_RGB565 || pix_fmt == PIX_FMT_RGB555) w = width * 2; - else if (pix_fmt == PIX_FMT_UYVY411) - w = width + width/2; - else if (pix_fmt == PIX_FMT_PAL8) - w = width; - else - w = width * (pf->depth * pf->nb_channels / 8); + else if (pix_fmt == PIX_FMT_UYVY411) + w = width + width/2; + else if (pix_fmt == PIX_FMT_PAL8) + w = width; + else + w = width * (pf->depth * pf->nb_channels / 8); - data_planes = 1; - h = height; + data_planes = 1; + h = height; } else { data_planes = pf->nb_channels; - w = (width*pf->depth + 7)/8; - h = height; + w = (width*pf->depth + 7)/8; + h = height; } for (i=0; i<data_planes; i++) { if (i == 1) { - w = width >> pf->x_chroma_shift; - h = height >> pf->y_chroma_shift; - } + w = width >> pf->x_chroma_shift; + h = height >> pf->y_chroma_shift; + } s = src->data[i]; - for(j=0; j<h; j++) { - memcpy(dest, s, w); - dest += w; - s += src->linesize[i]; - } + for(j=0; j<h; j++) { + memcpy(dest, s, w); + dest += w; + s += src->linesize[i]; + } } if (pf->pixel_type == FF_PIXEL_PALETTE) - memcpy((unsigned char *)(((size_t)dest + 3) & ~3), src->data[1], 256 * 4); + memcpy((unsigned char *)(((size_t)dest + 3) & ~3), src->data[1], 256 * 4); return size; } @@ -486,9 +486,9 @@ case PIX_FMT_RGB555: bits = 16; break; - case PIX_FMT_UYVY411: - bits = 12; - break; + case PIX_FMT_UYVY411: + bits = 12; + break; default: bits = pf->depth * pf->nb_channels; break; @@ -604,9 +604,9 @@ case PIX_FMT_RGB555: bits = 16; break; - case PIX_FMT_UYVY411: - bits = 12; - break; + case PIX_FMT_UYVY411: + bits = 12; + break; default: bits = pf->depth * pf->nb_channels; break; @@ -910,11 +910,11 @@ cr = cr1; for(w = width; w >= 4; w -= 4) { cb[0] = p[0]; - lum[0] = p[1]; + lum[0] = p[1]; lum[1] = p[2]; cr[0] = p[3]; - lum[2] = p[4]; - lum[3] = p[5]; + lum[2] = p[4]; + lum[3] = p[5]; p += 6; lum += 4; cb++; @@ -996,7 +996,7 @@ #define SCALEBITS 10 #define ONE_HALF (1 << (SCALEBITS - 1)) -#define FIX(x) ((int) ((x) * (1<<SCALEBITS) + 0.5)) +#define FIX(x) ((int) ((x) * (1<<SCALEBITS) + 0.5)) #define YUV_TO_RGB1_CCIR(cb1, cr1)\ {\ @@ -1046,7 +1046,7 @@ static inline int C_JPEG_TO_CCIR(int y) { y = (((y - 128) * FIX(112.0/127.0) + (ONE_HALF + (128 << SCALEBITS))) >> SCALEBITS); if (y < 16) - y = 16; + y = 16; return y; } @@ -1681,7 +1681,7 @@ typedef struct ConvertEntry { void (*convert)(AVPicture *dst, - const AVPicture *src, int width, int height); + const AVPicture *src, int width, int height); } ConvertEntry; /* Add each new convertion function in this table. In order to be able @@ -1721,7 +1721,7 @@ [PIX_FMT_RGBA32] = { .convert = yuv420p_to_rgba32 }, - [PIX_FMT_UYVY422] = { + [PIX_FMT_UYVY422] = { .convert = yuv420p_to_uyvy422, }, }, @@ -2224,7 +2224,7 @@ * @return ored mask of FF_ALPHA_xxx constants */ int img_get_alpha_info(const AVPicture *src, - int pix_fmt, int width, int height) + int pix_fmt, int width, int height) { PixFmtInfo *pf = &pix_fmt_info[pix_fmt]; int ret; @@ -2300,10 +2300,10 @@ /* filter parameters: [-1 4 2 4 -1] // 8 */ static void deinterlace_line(uint8_t *dst, - const uint8_t *lum_m4, const uint8_t *lum_m3, - const uint8_t *lum_m2, const uint8_t *lum_m1, - const uint8_t *lum, - int size) + const uint8_t *lum_m4, const uint8_t *lum_m3, + const uint8_t *lum_m2, const uint8_t *lum_m1, + const uint8_t *lum, + int size) { #ifndef HAVE_MMX uint8_t *cm = cropTbl + MAX_NEG_CROP; @@ -2421,7 +2421,7 @@ } static void deinterlace_bottom_field_inplace(uint8_t *src1, int src_wrap, - int width, int height) + int width, int height) { uint8_t *src_m1, *src_0, *src_p1, *src_p2; int y; @@ -2455,7 +2455,7 @@ if (pix_fmt != PIX_FMT_YUV420P && pix_fmt != PIX_FMT_YUV422P && pix_fmt != PIX_FMT_YUV444P && - pix_fmt != PIX_FMT_YUV411P) + pix_fmt != PIX_FMT_YUV411P) return -1; if ((width & 3) != 0 || (height & 3) != 0) return -1;
--- a/imgconvert_template.h Wed Dec 21 17:50:40 2005 +0000 +++ b/imgconvert_template.h Thu Dec 22 01:10:11 2005 +0000 @@ -821,7 +821,7 @@ #ifdef RGBA_IN static int glue(get_alpha_info_, RGB_NAME)(const AVPicture *src, - int width, int height) + int width, int height) { const unsigned char *p; int src_wrap, ret, x, y;
--- a/imgresample.c Wed Dec 21 17:50:40 2005 +0000 +++ b/imgresample.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,8 +64,8 @@ /* This function must be optimized */ static void h_resample_fast(uint8_t *dst, int dst_width, const uint8_t *src, - int src_width, int src_start, int src_incr, - int16_t *filters) + int src_width, int src_start, int src_incr, + int16_t *filters) { int src_pos, phase, sum, i; const uint8_t *s; @@ -108,7 +108,7 @@ /* This function must be optimized */ static void v_resample(uint8_t *dst, int dst_width, const uint8_t *src, - int wrap, int16_t *filter) + int wrap, int16_t *filter) { int sum, i; const uint8_t *s; @@ -167,7 +167,7 @@ /* XXX: do four pixels at a time */ static void h_resample_fast4_mmx(uint8_t *dst, int dst_width, - const uint8_t *src, int src_width, + const uint8_t *src, int src_width, int src_start, int src_incr, int16_t *filters) { int src_pos, phase; @@ -212,7 +212,7 @@ } static void v_resample4_mmx(uint8_t *dst, int dst_width, const uint8_t *src, - int wrap, int16_t *filter) + int wrap, int16_t *filter) { int sum, i, v; const uint8_t *s; @@ -277,18 +277,18 @@ #endif #ifdef HAVE_ALTIVEC -typedef union { +typedef union { vector unsigned char v; unsigned char c[16]; } vec_uc_t; -typedef union { +typedef union { vector signed short v; signed short s[8]; } vec_ss_t; void v_resample16_altivec(uint8_t *dst, int dst_width, const uint8_t *src, - int wrap, int16_t *filter) + int wrap, int16_t *filter) { int sum, i; const uint8_t *s; @@ -405,7 +405,7 @@ /* slow version to handle limit cases. Does not need optimisation */ static void h_resample_slow(uint8_t *dst, int dst_width, - const uint8_t *src, int src_width, + const uint8_t *src, int src_width, int src_start, int src_incr, int16_t *filters) { int src_pos, phase, sum, j, v, i; @@ -441,8 +441,8 @@ } static void h_resample(uint8_t *dst, int dst_width, const uint8_t *src, - int src_width, int src_start, int src_incr, - int16_t *filters) + int src_width, int src_start, int src_incr, + int16_t *filters) { int n, src_end; @@ -559,7 +559,7 @@ ImgReSampleContext *s; if (!owidth || !oheight || !iwidth || !iheight) - return NULL; + return NULL; s = av_mallocz(sizeof(ImgReSampleContext)); if (!s)
--- a/indeo3.c Wed Dec 21 17:50:40 2005 +0000 +++ b/indeo3.c Thu Dec 22 01:10:11 2005 +0000 @@ -70,13 +70,13 @@ for (i=0; i < 128; ++i) { s->ModPred[i+0*128] = (i > 126) ? 254 : 2*((i + 1) - ((i + 1) % 2)); s->ModPred[i+1*128] = (i == 7) ? 20 : ((i == 119 || i == 120) - ? 236 : 2*((i + 2) - ((i + 1) % 3))); + ? 236 : 2*((i + 2) - ((i + 1) % 3))); s->ModPred[i+2*128] = (i > 125) ? 248 : 2*((i + 2) - ((i + 2) % 4)); - s->ModPred[i+3*128] = 2*((i + 1) - ((i - 3) % 5)); + s->ModPred[i+3*128] = 2*((i + 1) - ((i - 3) % 5)); s->ModPred[i+4*128] = (i == 8) ? 20 : 2*((i + 1) - ((i - 3) % 6)); - s->ModPred[i+5*128] = 2*((i + 4) - ((i + 3) % 7)); + s->ModPred[i+5*128] = 2*((i + 4) - ((i + 3) % 7)); s->ModPred[i+6*128] = (i > 123) ? 240 : 2*((i + 4) - ((i + 4) % 8)); - s->ModPred[i+7*128] = 2*((i + 5) - ((i + 4) % 9)); + s->ModPred[i+7*128] = 2*((i + 5) - ((i + 4) % 9)); } s->corrector_type = (unsigned short *) av_malloc (24 * 256 * sizeof(unsigned short)); @@ -84,8 +84,8 @@ for (i=0; i < 24; ++i) { for (j=0; j < 256; ++j) { s->corrector_type[i*256+j] = (j < corrector_type_0[i]) - ? 1 : ((j < 248 || (i == 16 && j == 248)) - ? 0 : corrector_type_2[j - 248]); + ? 1 : ((j < 248 || (i == 16 && j == 248)) + ? 0 : corrector_type_2[j - 248]); } } }
--- a/indeo3data.h Wed Dec 21 17:50:40 2005 +0000 +++ b/indeo3data.h Thu Dec 22 01:10:11 2005 +0000 @@ -1,2315 +1,2315 @@ static const uint32_t correction[] = { - 0x00000000, 0x00000202, 0xfffffdfe, 0x000002ff, 0xfffffd01, 0xffffff03, 0x000000fd, 0x00000404, - 0xfffffbfc, 0x00000501, 0xfffffaff, 0x00000105, 0xfffffefb, 0x000003fc, 0xfffffc04, 0x000005fe, - 0xfffffa02, 0xfffffe06, 0x000001fa, 0x00000904, 0xfffff6fc, 0x00000409, 0xfffffbf7, 0x00000909, - 0xfffff6f7, 0x00000a01, 0xfffff5ff, 0x0000010a, 0xfffffef6, 0x000007fb, 0xfffff805, 0xfffffb08, - 0x000004f8, 0x00000f09, 0xfffff0f7, 0x0000090f, 0xfffff6f1, 0x00000bfd, 0xfffff403, 0xfffffd0c, - 0x000002f4, 0x00001004, 0xffffeffc, 0x00000410, 0xfffffbf0, 0x00001010, 0xffffeff0, 0x00001200, - 0xffffee00, 0x00000012, 0xffffffee, 0x00000bf4, 0xfffff40c, 0x00000ff7, 0xfffff009, 0xfffff710, - 0x000008f0, 0x00001b0b, 0xffffe4f5, 0x00000b1b, 0xfffff4e5, 0x00001c13, 0xffffe3ed, 0x0000131c, - 0xffffece4, 0x000015fa, 0xffffea06, 0xfffffa16, 0x000005ea, 0x00001d04, 0xffffe2fc, 0x0000041d, - 0xfffffbe3, 0x00001e1e, 0xffffe1e2, 0x000020fe, 0xffffdf02, 0xfffffe21, 0x000001df, 0x000016ee, - 0xffffe912, 0xffffee17, 0x000011e9, 0x00001df1, 0xffffe20f, 0xfffff11e, 0x00000ee2, 0x00002e16, - 0xffffd1ea, 0x0000162e, 0xffffe9d2, 0x00002f0d, 0xffffd0f3, 0x00000d2f, 0xfffff2d1, 0x00003123, - 0xffffcedd, 0x00002331, 0xffffdccf, 0x000028f5, 0xffffd70b, 0xfffff529, 0x00000ad7, 0x00003304, - 0xffffccfc, 0x00000433, 0xfffffbcd, 0x00003636, 0xffffc9ca, 0x000021de, 0xffffde22, 0x000029e3, - 0xffffd61d, 0xffffe32a, 0x00001cd6, 0x00003bfa, 0xffffc406, 0xfffffa3c, 0x000005c4, 0x00004c1b, - 0xffffb3e5, 0x00001b4c, 0xffffe4b4, 0x00004d2b, 0xffffb2d5, 0x00002b4d, 0xffffd4b3, 0x000036e8, - 0xffffc918, 0xffffe837, 0x000017c9, 0x00004f0e, 0xffffb0f2, 0x00000e4f, 0xfffff1b1, 0x0000533f, - 0xffffacc1, 0x00003f53, 0xffffc0ad, 0x000049ec, 0xffffb614, 0xffffec4a, 0x000013b6, 0x00005802, - 0xffffa7fe, 0x00000258, 0xfffffda8, 0x00005d5d, 0xffffa2a3, 0x00003ccc, 0xffffc334, 0xffffcc3d, - 0x000033c3, 0x00007834, 0xffff87cc, 0x00003478, 0xffffcb88, 0x00004ad3, 0xffffb52d, 0xffffd34b, - 0x00002cb5, 0x00007d4b, 0xffff82b5, 0x00004b7d, 0xffffb483, 0x00007a21, 0xffff85df, 0x0000217a, - 0xffffde86, 0x000066f3, 0xffff990d, 0xfffff367, 0x00000c99, 0x00005fd8, 0xffffa028, 0xffffd860, - 0x000027a0, 0x00007ede, 0xffff8122, 0xffffde7f, 0x00002181, 0x000058a7, 0xffffa759, 0x000068b2, - 0xffff974e, 0xffffb269, 0x00004d97, 0x00000c0c, 0xfffff3f4, 0x00001717, 0xffffe8e9, 0x00002a2a, - 0xffffd5d6, 0x00004949, 0xffffb6b7, 0x00000000, 0x02020000, 0xfdfe0000, 0x02ff0000, 0xfd010000, - 0xff030000, 0x00fd0000, 0x00000202, 0x02020202, 0xfdfe0202, 0x02ff0202, 0xfd010202, 0xff030202, - 0x00fd0202, 0xfffffdfe, 0x0201fdfe, 0xfdfdfdfe, 0x02fefdfe, 0xfd00fdfe, 0xff02fdfe, 0x00fcfdfe, - 0x000002ff, 0x020202ff, 0xfdfe02ff, 0x02ff02ff, 0xfd0102ff, 0xff0302ff, 0x00fd02ff, 0xfffffd01, - 0x0201fd01, 0xfdfdfd01, 0x02fefd01, 0xfd00fd01, 0xff02fd01, 0x00fcfd01, 0xffffff03, 0x0201ff03, - 0xfdfdff03, 0x02feff03, 0xfd00ff03, 0xff02ff03, 0x00fcff03, 0x000000fd, 0x020200fd, 0xfdfe00fd, - 0x02ff00fd, 0xfd0100fd, 0xff0300fd, 0x00fd00fd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000303, 0xfffffcfd, 0x000003ff, 0xfffffc01, 0xffffff04, 0x000000fc, 0x00000707, - 0xfffff8f9, 0x00000802, 0xfffff7fe, 0x00000208, 0xfffffdf8, 0x000008fe, 0xfffff702, 0xfffffe09, - 0x000001f7, 0x000005fa, 0xfffffa06, 0x00000d06, 0xfffff2fa, 0x0000060d, 0xfffff9f3, 0x00000d0d, - 0xfffff2f3, 0x00000e01, 0xfffff1ff, 0x0000010e, 0xfffffef2, 0x00000bf8, 0xfffff408, 0xfffff80c, - 0x000007f4, 0x0000170e, 0xffffe8f2, 0x00000e17, 0xfffff1e9, 0x000011fb, 0xffffee05, 0xfffffb12, - 0x000004ee, 0x00001806, 0xffffe7fa, 0x00000618, 0xfffff9e8, 0x00001818, 0xffffe7e8, 0x00001aff, - 0xffffe501, 0xffffff1b, 0x000000e5, 0x000010ef, 0xffffef11, 0x000016f3, 0xffffe90d, 0xfffff317, - 0x00000ce9, 0x00002810, 0xffffd7f0, 0x00001028, 0xffffefd8, 0x0000291c, 0xffffd6e4, 0x00001c29, - 0xffffe3d7, 0x000020f7, 0xffffdf09, 0xfffff721, 0x000008df, 0x00002b06, 0xffffd4fa, 0x0000062b, - 0xfffff9d5, 0x00002e2e, 0xffffd1d2, 0x000031fc, 0xffffce04, 0xfffffc32, 0x000003ce, 0x000021e5, - 0xffffde1b, 0xffffe522, 0x00001ade, 0x00002cea, 0xffffd316, 0xffffea2d, 0x000015d3, 0x00004522, - 0xffffbade, 0x00002245, 0xffffddbb, 0x00004613, 0xffffb9ed, 0x00001346, 0xffffecba, 0x00004935, - 0xffffb6cb, 0x00003549, 0xffffcab7, 0x00003def, 0xffffc211, 0xffffef3e, 0x000010c2, 0x00004d05, - 0xffffb2fb, 0x0000054d, 0xfffffab3, 0x00005252, 0xffffadae, 0x000032cd, 0xffffcd33, 0x00003fd5, - 0xffffc02b, 0xffffd540, 0x00002ac0, 0x000059f6, 0xffffa60a, 0xfffff65a, 0x000009a6, 0x00007229, - 0xffff8dd7, 0x00002972, 0xffffd68e, 0x00007440, 0xffff8bc0, 0x00004074, 0xffffbf8c, 0x000051db, - 0xffffae25, 0xffffdb52, 0x000024ae, 0x00007716, 0xffff88ea, 0x00001677, 0xffffe989, 0x00007c5f, - 0xffff83a1, 0x00005f7c, 0xffffa084, 0x00006ee2, 0xffff911e, 0xffffe26f, 0x00001d91, 0x00005bb2, - 0xffffa44e, 0xffffb25c, 0x00004da4, 0x000070bc, 0xffff8f44, 0xffffbc71, 0x0000438f, 0x00001212, - 0xffffedee, 0x00002222, 0xffffddde, 0x00003f3f, 0xffffc0c1, 0x00006d6d, 0xffff9293, 0x00000000, - 0x03030000, 0xfcfd0000, 0x03ff0000, 0xfc010000, 0xff040000, 0x00fc0000, 0x07070000, 0xf8f90000, - 0x00000303, 0x03030303, 0xfcfd0303, 0x03ff0303, 0xfc010303, 0xff040303, 0x00fc0303, 0x07070303, - 0xf8f90303, 0xfffffcfd, 0x0302fcfd, 0xfcfcfcfd, 0x03fefcfd, 0xfc00fcfd, 0xff03fcfd, 0x00fbfcfd, - 0x0706fcfd, 0xf8f8fcfd, 0x000003ff, 0x030303ff, 0xfcfd03ff, 0x03ff03ff, 0xfc0103ff, 0xff0403ff, - 0x00fc03ff, 0x070703ff, 0xf8f903ff, 0xfffffc01, 0x0302fc01, 0xfcfcfc01, 0x03fefc01, 0xfc00fc01, - 0xff03fc01, 0x00fbfc01, 0x0706fc01, 0xf8f8fc01, 0xffffff04, 0x0302ff04, 0xfcfcff04, 0x03feff04, - 0xfc00ff04, 0xff03ff04, 0x00fbff04, 0x0706ff04, 0xf8f8ff04, 0x000000fc, 0x030300fc, 0xfcfd00fc, - 0x03ff00fc, 0xfc0100fc, 0xff0400fc, 0x00fc00fc, 0x070700fc, 0xf8f900fc, 0x00000707, 0x03030707, - 0xfcfd0707, 0x03ff0707, 0xfc010707, 0xff040707, 0x00fc0707, 0x07070707, 0xf8f90707, 0xfffff8f9, - 0x0302f8f9, 0xfcfcf8f9, 0x03fef8f9, 0xfc00f8f9, 0xff03f8f9, 0x00fbf8f9, 0x0706f8f9, 0xf8f8f8f9, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000404, 0xfffffbfc, 0x000004ff, 0xfffffb01, 0xffffff05, 0x000000fb, 0x00000a03, - 0xfffff5fd, 0x0000030a, 0xfffffcf6, 0x00000909, 0xfffff6f7, 0x000006f9, 0xfffff907, 0x00000bfd, - 0xfffff403, 0xfffffd0c, 0x000002f4, 0x00001108, 0xffffeef8, 0x00000811, 0xfffff7ef, 0x00001111, - 0xffffeeef, 0x00001301, 0xffffecff, 0x00000113, 0xfffffeed, 0x00000ff5, 0xfffff00b, 0xfffff510, - 0x00000af0, 0x000016fa, 0xffffe906, 0xfffffa17, 0x000005e9, 0x00001f12, 0xffffe0ee, 0x0000121f, - 0xffffede1, 0x00002008, 0xffffdff8, 0x00000820, 0xfffff7e0, 0x00002121, 0xffffdedf, 0x000023ff, - 0xffffdc01, 0xffffff24, 0x000000dc, 0x000016e9, 0xffffe917, 0x00001eef, 0xffffe111, 0xffffef1f, - 0x000010e1, 0x00003615, 0xffffc9eb, 0x00001536, 0xffffeaca, 0x00003725, 0xffffc8db, 0x00002537, - 0xffffdac9, 0x00002bf4, 0xffffd40c, 0xfffff42c, 0x00000bd4, 0x00003908, 0xffffc6f8, 0x00000839, - 0xfffff7c7, 0x00003d3d, 0xffffc2c3, 0x000041fb, 0xffffbe05, 0xfffffb42, 0x000004be, 0x00002cdc, - 0xffffd324, 0xffffdc2d, 0x000023d3, 0x00003be3, 0xffffc41d, 0xffffe33c, 0x00001cc4, 0x00005c2d, - 0xffffa3d3, 0x00002d5c, 0xffffd2a4, 0x00005d19, 0xffffa2e7, 0x0000195d, 0xffffe6a3, 0x00006147, - 0xffff9eb9, 0x00004761, 0xffffb89f, 0x000052ea, 0xffffad16, 0xffffea53, 0x000015ad, 0x00006607, - 0xffff99f9, 0x00000766, 0xfffff89a, 0x00006d6d, 0xffff9293, 0x000043bc, 0xffffbc44, 0x000054c7, - 0xffffab39, 0xffffc755, 0x000038ab, 0x000077f3, 0xffff880d, 0xfffff378, 0x00000c88, 0x00006dcf, - 0xffff9231, 0xffffcf6e, 0x00003092, 0x00007a98, 0xffff8568, 0xffff987b, 0x00006785, 0x00001818, - 0xffffe7e8, 0x00002e2e, 0xffffd1d2, 0x00005454, 0xffffabac, 0x00000000, 0x04040000, 0xfbfc0000, - 0x04ff0000, 0xfb010000, 0xff050000, 0x00fb0000, 0x0a030000, 0xf5fd0000, 0x030a0000, 0x00000404, - 0x04040404, 0xfbfc0404, 0x04ff0404, 0xfb010404, 0xff050404, 0x00fb0404, 0x0a030404, 0xf5fd0404, - 0x030a0404, 0xfffffbfc, 0x0403fbfc, 0xfbfbfbfc, 0x04fefbfc, 0xfb00fbfc, 0xff04fbfc, 0x00fafbfc, - 0x0a02fbfc, 0xf5fcfbfc, 0x0309fbfc, 0x000004ff, 0x040404ff, 0xfbfc04ff, 0x04ff04ff, 0xfb0104ff, - 0xff0504ff, 0x00fb04ff, 0x0a0304ff, 0xf5fd04ff, 0x030a04ff, 0xfffffb01, 0x0403fb01, 0xfbfbfb01, - 0x04fefb01, 0xfb00fb01, 0xff04fb01, 0x00fafb01, 0x0a02fb01, 0xf5fcfb01, 0x0309fb01, 0xffffff05, - 0x0403ff05, 0xfbfbff05, 0x04feff05, 0xfb00ff05, 0xff04ff05, 0x00faff05, 0x0a02ff05, 0xf5fcff05, - 0x0309ff05, 0x000000fb, 0x040400fb, 0xfbfc00fb, 0x04ff00fb, 0xfb0100fb, 0xff0500fb, 0x00fb00fb, - 0x0a0300fb, 0xf5fd00fb, 0x030a00fb, 0x00000a03, 0x04040a03, 0xfbfc0a03, 0x04ff0a03, 0xfb010a03, - 0xff050a03, 0x00fb0a03, 0x0a030a03, 0xf5fd0a03, 0x030a0a03, 0xfffff5fd, 0x0403f5fd, 0xfbfbf5fd, - 0x04fef5fd, 0xfb00f5fd, 0xff04f5fd, 0x00faf5fd, 0x0a02f5fd, 0xf5fcf5fd, 0x0309f5fd, 0x0000030a, - 0x0404030a, 0xfbfc030a, 0x04ff030a, 0xfb01030a, 0xff05030a, 0x00fb030a, 0x0a03030a, 0xf5fd030a, - 0x030a030a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000505, 0xfffffafb, 0x000006fe, 0xfffff902, 0xfffffe07, 0x000001f9, 0x00000b0b, - 0xfffff4f5, 0x00000d03, 0xfffff2fd, 0x0000030d, 0xfffffcf3, 0x000008f7, 0xfffff709, 0x00000efc, - 0xfffff104, 0xfffffc0f, 0x000003f1, 0x0000160b, 0xffffe9f5, 0x00000b16, 0xfffff4ea, 0x00001515, - 0xffffeaeb, 0x00001802, 0xffffe7fe, 0x00000218, 0xfffffde8, 0x000013f2, 0xffffec0e, 0xfffff214, - 0x00000dec, 0x00002617, 0xffffd9e9, 0x00001726, 0xffffe8da, 0x00001cf8, 0xffffe308, 0xfffff81d, - 0x000007e3, 0x0000270b, 0xffffd8f5, 0x00000b27, 0xfffff4d9, 0x00002929, 0xffffd6d7, 0x00002cff, - 0xffffd301, 0xffffff2d, 0x000000d3, 0x00001ce3, 0xffffe31d, 0x000026ea, 0xffffd916, 0xffffea27, - 0x000015d9, 0x0000431b, 0xffffbce5, 0x00001b43, 0xffffe4bd, 0x0000452f, 0xffffbad1, 0x00002f45, - 0xffffd0bb, 0x000037f1, 0xffffc80f, 0xfffff138, 0x00000ec8, 0x0000470b, 0xffffb8f5, 0x00000b47, - 0xfffff4b9, 0x00004c4c, 0xffffb3b4, 0x000052fa, 0xffffad06, 0xfffffa53, 0x000005ad, 0x000038d3, - 0xffffc72d, 0xffffd339, 0x00002cc7, 0x00004adc, 0xffffb524, 0xffffdc4b, 0x000023b5, 0x00007338, - 0xffff8cc8, 0x00003873, 0xffffc78d, 0x0000751f, 0xffff8ae1, 0x00001f75, 0xffffe08b, 0x00007a58, - 0xffff85a8, 0x0000587a, 0xffffa786, 0x000067e4, 0xffff981c, 0xffffe468, 0x00001b98, 0x000054ab, - 0xffffab55, 0x000069b8, 0xffff9648, 0xffffb86a, 0x00004796, 0x00001e1e, 0xffffe1e2, 0x00003a3a, - 0xffffc5c6, 0x00006969, 0xffff9697, 0x00000000, 0x05050000, 0xfafb0000, 0x06fe0000, 0xf9020000, - 0xfe070000, 0x01f90000, 0x0b0b0000, 0xf4f50000, 0x0d030000, 0xf2fd0000, 0x00000505, 0x05050505, - 0xfafb0505, 0x06fe0505, 0xf9020505, 0xfe070505, 0x01f90505, 0x0b0b0505, 0xf4f50505, 0x0d030505, - 0xf2fd0505, 0xfffffafb, 0x0504fafb, 0xfafafafb, 0x06fdfafb, 0xf901fafb, 0xfe06fafb, 0x01f8fafb, - 0x0b0afafb, 0xf4f4fafb, 0x0d02fafb, 0xf2fcfafb, 0x000006fe, 0x050506fe, 0xfafb06fe, 0x06fe06fe, - 0xf90206fe, 0xfe0706fe, 0x01f906fe, 0x0b0b06fe, 0xf4f506fe, 0x0d0306fe, 0xf2fd06fe, 0xfffff902, - 0x0504f902, 0xfafaf902, 0x06fdf902, 0xf901f902, 0xfe06f902, 0x01f8f902, 0x0b0af902, 0xf4f4f902, - 0x0d02f902, 0xf2fcf902, 0xfffffe07, 0x0504fe07, 0xfafafe07, 0x06fdfe07, 0xf901fe07, 0xfe06fe07, - 0x01f8fe07, 0x0b0afe07, 0xf4f4fe07, 0x0d02fe07, 0xf2fcfe07, 0x000001f9, 0x050501f9, 0xfafb01f9, - 0x06fe01f9, 0xf90201f9, 0xfe0701f9, 0x01f901f9, 0x0b0b01f9, 0xf4f501f9, 0x0d0301f9, 0xf2fd01f9, - 0x00000b0b, 0x05050b0b, 0xfafb0b0b, 0x06fe0b0b, 0xf9020b0b, 0xfe070b0b, 0x01f90b0b, 0x0b0b0b0b, - 0xf4f50b0b, 0x0d030b0b, 0xf2fd0b0b, 0xfffff4f5, 0x0504f4f5, 0xfafaf4f5, 0x06fdf4f5, 0xf901f4f5, - 0xfe06f4f5, 0x01f8f4f5, 0x0b0af4f5, 0xf4f4f4f5, 0x0d02f4f5, 0xf2fcf4f5, 0x00000d03, 0x05050d03, - 0xfafb0d03, 0x06fe0d03, 0xf9020d03, 0xfe070d03, 0x01f90d03, 0x0b0b0d03, 0xf4f50d03, 0x0d030d03, - 0xf2fd0d03, 0xfffff2fd, 0x0504f2fd, 0xfafaf2fd, 0x06fdf2fd, 0xf901f2fd, 0xfe06f2fd, 0x01f8f2fd, - 0x0b0af2fd, 0xf4f4f2fd, 0x0d02f2fd, 0xf2fcf2fd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000606, 0xfffff9fa, 0x000007fe, 0xfffff802, 0xfffffe08, 0x000001f8, 0x00000d0d, - 0xfffff2f3, 0x00000f04, 0xfffff0fc, 0x0000040f, 0xfffffbf1, 0x00000af5, 0xfffff50b, 0x000011fb, - 0xffffee05, 0xfffffb12, 0x000004ee, 0x00001a0d, 0xffffe5f3, 0x00000d1a, 0xfffff2e6, 0x00001a1a, - 0xffffe5e6, 0x00001d02, 0xffffe2fe, 0x0000021d, 0xfffffde3, 0x000017f0, 0xffffe810, 0xfffff018, - 0x00000fe8, 0x00002e1c, 0xffffd1e4, 0x00001c2e, 0xffffe3d2, 0x000022f7, 0xffffdd09, 0xfffff723, - 0x000008dd, 0x00002f0d, 0xffffd0f3, 0x00000d2f, 0xfffff2d1, 0x00003131, 0xffffcecf, 0x000035ff, - 0xffffca01, 0xffffff36, 0x000000ca, 0x000022dd, 0xffffdd23, 0x00002ee6, 0xffffd11a, 0xffffe62f, - 0x000019d1, 0x00005120, 0xffffaee0, 0x00002051, 0xffffdfaf, 0x00005338, 0xffffacc8, 0x00003853, - 0xffffc7ad, 0x000042ee, 0xffffbd12, 0xffffee43, 0x000011bd, 0x0000560d, 0xffffa9f3, 0x00000d56, - 0xfffff2aa, 0x00005b5b, 0xffffa4a5, 0x000062f9, 0xffff9d07, 0xfffff963, 0x0000069d, 0x000043ca, - 0xffffbc36, 0xffffca44, 0x000035bc, 0x000059d4, 0xffffa62c, 0xffffd45a, 0x00002ba6, 0x00007bdf, - 0xffff8421, 0xffffdf7c, 0x00002084, 0x00006699, 0xffff9967, 0x00007eaa, 0xffff8156, 0xffffaa7f, - 0x00005581, 0x00002525, 0xffffdadb, 0x00004545, 0xffffbabb, 0x00000000, 0x06060000, 0xf9fa0000, - 0x07fe0000, 0xf8020000, 0xfe080000, 0x01f80000, 0x0d0d0000, 0xf2f30000, 0x0f040000, 0xf0fc0000, - 0x040f0000, 0x00000606, 0x06060606, 0xf9fa0606, 0x07fe0606, 0xf8020606, 0xfe080606, 0x01f80606, - 0x0d0d0606, 0xf2f30606, 0x0f040606, 0xf0fc0606, 0x040f0606, 0xfffff9fa, 0x0605f9fa, 0xf9f9f9fa, - 0x07fdf9fa, 0xf801f9fa, 0xfe07f9fa, 0x01f7f9fa, 0x0d0cf9fa, 0xf2f2f9fa, 0x0f03f9fa, 0xf0fbf9fa, - 0x040ef9fa, 0x000007fe, 0x060607fe, 0xf9fa07fe, 0x07fe07fe, 0xf80207fe, 0xfe0807fe, 0x01f807fe, - 0x0d0d07fe, 0xf2f307fe, 0x0f0407fe, 0xf0fc07fe, 0x040f07fe, 0xfffff802, 0x0605f802, 0xf9f9f802, - 0x07fdf802, 0xf801f802, 0xfe07f802, 0x01f7f802, 0x0d0cf802, 0xf2f2f802, 0x0f03f802, 0xf0fbf802, - 0x040ef802, 0xfffffe08, 0x0605fe08, 0xf9f9fe08, 0x07fdfe08, 0xf801fe08, 0xfe07fe08, 0x01f7fe08, - 0x0d0cfe08, 0xf2f2fe08, 0x0f03fe08, 0xf0fbfe08, 0x040efe08, 0x000001f8, 0x060601f8, 0xf9fa01f8, - 0x07fe01f8, 0xf80201f8, 0xfe0801f8, 0x01f801f8, 0x0d0d01f8, 0xf2f301f8, 0x0f0401f8, 0xf0fc01f8, - 0x040f01f8, 0x00000d0d, 0x06060d0d, 0xf9fa0d0d, 0x07fe0d0d, 0xf8020d0d, 0xfe080d0d, 0x01f80d0d, - 0x0d0d0d0d, 0xf2f30d0d, 0x0f040d0d, 0xf0fc0d0d, 0x040f0d0d, 0xfffff2f3, 0x0605f2f3, 0xf9f9f2f3, - 0x07fdf2f3, 0xf801f2f3, 0xfe07f2f3, 0x01f7f2f3, 0x0d0cf2f3, 0xf2f2f2f3, 0x0f03f2f3, 0xf0fbf2f3, - 0x040ef2f3, 0x00000f04, 0x06060f04, 0xf9fa0f04, 0x07fe0f04, 0xf8020f04, 0xfe080f04, 0x01f80f04, - 0x0d0d0f04, 0xf2f30f04, 0x0f040f04, 0xf0fc0f04, 0x040f0f04, 0xfffff0fc, 0x0605f0fc, 0xf9f9f0fc, - 0x07fdf0fc, 0xf801f0fc, 0xfe07f0fc, 0x01f7f0fc, 0x0d0cf0fc, 0xf2f2f0fc, 0x0f03f0fc, 0xf0fbf0fc, - 0x040ef0fc, 0x0000040f, 0x0606040f, 0xf9fa040f, 0x07fe040f, 0xf802040f, 0xfe08040f, 0x01f8040f, - 0x0d0d040f, 0xf2f3040f, 0x0f04040f, 0xf0fc040f, 0x040f040f, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000707, 0xfffff8f9, 0x000009fd, 0xfffff603, 0xfffffd0a, 0x000002f6, 0x00001010, - 0xffffeff0, 0x00001205, 0xffffedfb, 0x00000512, 0xfffffaee, 0x00000cf3, 0xfffff30d, 0x000014fa, - 0xffffeb06, 0xfffffa15, 0x000005eb, 0x00001e0f, 0xffffe1f1, 0x00000f1e, 0xfffff0e2, 0x00001e1e, - 0xffffe1e2, 0x00002202, 0xffffddfe, 0x00000222, 0xfffffdde, 0x00001bed, 0xffffe413, 0xffffed1c, - 0x000012e4, 0x00003620, 0xffffc9e0, 0x00002036, 0xffffdfca, 0x000028f5, 0xffffd70b, 0xfffff529, - 0x00000ad7, 0x0000370f, 0xffffc8f1, 0x00000f37, 0xfffff0c9, 0x00003939, 0xffffc6c7, 0x00003eff, - 0xffffc101, 0xffffff3f, 0x000000c1, 0x000027d8, 0xffffd828, 0x000036e2, 0xffffc91e, 0xffffe237, - 0x00001dc9, 0x00005e25, 0xffffa1db, 0x0000255e, 0xffffdaa2, 0x00006041, 0xffff9fbf, 0x00004160, - 0xffffbea0, 0x00004deb, 0xffffb215, 0xffffeb4e, 0x000014b2, 0x0000640f, 0xffff9bf1, 0x00000f64, - 0xfffff09c, 0x00006a6a, 0xffff9596, 0x000073f8, 0xffff8c08, 0xfffff874, 0x0000078c, 0x00004ec1, - 0xffffb13f, 0xffffc14f, 0x00003eb1, 0x000068cd, 0xffff9733, 0xffffcd69, 0x00003297, 0x00007788, - 0xffff8878, 0x00002b2b, 0xffffd4d5, 0x00005050, 0xffffafb0, 0x00000000, 0x07070000, 0xf8f90000, - 0x09fd0000, 0xf6030000, 0xfd0a0000, 0x02f60000, 0x10100000, 0xeff00000, 0x12050000, 0xedfb0000, - 0x05120000, 0x00000707, 0x07070707, 0xf8f90707, 0x09fd0707, 0xf6030707, 0xfd0a0707, 0x02f60707, - 0x10100707, 0xeff00707, 0x12050707, 0xedfb0707, 0x05120707, 0xfffff8f9, 0x0706f8f9, 0xf8f8f8f9, - 0x09fcf8f9, 0xf602f8f9, 0xfd09f8f9, 0x02f5f8f9, 0x100ff8f9, 0xefeff8f9, 0x1204f8f9, 0xedfaf8f9, - 0x0511f8f9, 0x000009fd, 0x070709fd, 0xf8f909fd, 0x09fd09fd, 0xf60309fd, 0xfd0a09fd, 0x02f609fd, - 0x101009fd, 0xeff009fd, 0x120509fd, 0xedfb09fd, 0x051209fd, 0xfffff603, 0x0706f603, 0xf8f8f603, - 0x09fcf603, 0xf602f603, 0xfd09f603, 0x02f5f603, 0x100ff603, 0xefeff603, 0x1204f603, 0xedfaf603, - 0x0511f603, 0xfffffd0a, 0x0706fd0a, 0xf8f8fd0a, 0x09fcfd0a, 0xf602fd0a, 0xfd09fd0a, 0x02f5fd0a, - 0x100ffd0a, 0xefeffd0a, 0x1204fd0a, 0xedfafd0a, 0x0511fd0a, 0x000002f6, 0x070702f6, 0xf8f902f6, - 0x09fd02f6, 0xf60302f6, 0xfd0a02f6, 0x02f602f6, 0x101002f6, 0xeff002f6, 0x120502f6, 0xedfb02f6, - 0x051202f6, 0x00001010, 0x07071010, 0xf8f91010, 0x09fd1010, 0xf6031010, 0xfd0a1010, 0x02f61010, - 0x10101010, 0xeff01010, 0x12051010, 0xedfb1010, 0x05121010, 0xffffeff0, 0x0706eff0, 0xf8f8eff0, - 0x09fceff0, 0xf602eff0, 0xfd09eff0, 0x02f5eff0, 0x100feff0, 0xefefeff0, 0x1204eff0, 0xedfaeff0, - 0x0511eff0, 0x00001205, 0x07071205, 0xf8f91205, 0x09fd1205, 0xf6031205, 0xfd0a1205, 0x02f61205, - 0x10101205, 0xeff01205, 0x12051205, 0xedfb1205, 0x05121205, 0xffffedfb, 0x0706edfb, 0xf8f8edfb, - 0x09fcedfb, 0xf602edfb, 0xfd09edfb, 0x02f5edfb, 0x100fedfb, 0xefefedfb, 0x1204edfb, 0xedfaedfb, - 0x0511edfb, 0x00000512, 0x07070512, 0xf8f90512, 0x09fd0512, 0xf6030512, 0xfd0a0512, 0x02f60512, - 0x10100512, 0xeff00512, 0x12050512, 0xedfb0512, 0x05120512, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000808, 0xfffff7f8, 0x00000afd, 0xfffff503, 0xfffffd0b, 0x000002f5, 0x00001212, - 0xffffedee, 0x00001405, 0xffffebfb, 0x00000514, 0xfffffaec, 0x00000ef1, 0xfffff10f, 0x000017f9, - 0xffffe807, 0xfffff918, 0x000006e8, 0x00002311, 0xffffdcef, 0x00001123, 0xffffeedd, 0x00002222, - 0xffffddde, 0x00002603, 0xffffd9fd, 0x00000326, 0xfffffcda, 0x00001fea, 0xffffe016, 0xffffea20, - 0x000015e0, 0x00003d25, 0xffffc2db, 0x0000253d, 0xffffdac3, 0x00002ef3, 0xffffd10d, 0xfffff32f, - 0x00000cd1, 0x00003f11, 0xffffc0ef, 0x0000113f, 0xffffeec1, 0x00004141, 0xffffbebf, 0x000047ff, - 0xffffb801, 0xffffff48, 0x000000b8, 0x00002dd2, 0xffffd22e, 0x00003edd, 0xffffc123, 0xffffdd3f, - 0x000022c1, 0x00006b2b, 0xffff94d5, 0x00002b6b, 0xffffd495, 0x00006e4b, 0xffff91b5, 0x00004b6e, - 0xffffb492, 0x000058e8, 0xffffa718, 0xffffe859, 0x000017a7, 0x00007211, 0xffff8def, 0x00001172, - 0xffffee8e, 0x00007979, 0xffff8687, 0x00005ab8, 0xffffa548, 0xffffb85b, 0x000047a5, 0x000077c6, - 0xffff883a, 0xffffc678, 0x00003988, 0x00003131, 0xffffcecf, 0x00005c5c, 0xffffa3a4, 0x00000000, - 0x08080000, 0xf7f80000, 0x0afd0000, 0xf5030000, 0xfd0b0000, 0x02f50000, 0x12120000, 0xedee0000, - 0x14050000, 0xebfb0000, 0x05140000, 0x00000808, 0x08080808, 0xf7f80808, 0x0afd0808, 0xf5030808, - 0xfd0b0808, 0x02f50808, 0x12120808, 0xedee0808, 0x14050808, 0xebfb0808, 0x05140808, 0xfffff7f8, - 0x0807f7f8, 0xf7f7f7f8, 0x0afcf7f8, 0xf502f7f8, 0xfd0af7f8, 0x02f4f7f8, 0x1211f7f8, 0xededf7f8, - 0x1404f7f8, 0xebfaf7f8, 0x0513f7f8, 0x00000afd, 0x08080afd, 0xf7f80afd, 0x0afd0afd, 0xf5030afd, - 0xfd0b0afd, 0x02f50afd, 0x12120afd, 0xedee0afd, 0x14050afd, 0xebfb0afd, 0x05140afd, 0xfffff503, - 0x0807f503, 0xf7f7f503, 0x0afcf503, 0xf502f503, 0xfd0af503, 0x02f4f503, 0x1211f503, 0xededf503, - 0x1404f503, 0xebfaf503, 0x0513f503, 0xfffffd0b, 0x0807fd0b, 0xf7f7fd0b, 0x0afcfd0b, 0xf502fd0b, - 0xfd0afd0b, 0x02f4fd0b, 0x1211fd0b, 0xededfd0b, 0x1404fd0b, 0xebfafd0b, 0x0513fd0b, 0x000002f5, - 0x080802f5, 0xf7f802f5, 0x0afd02f5, 0xf50302f5, 0xfd0b02f5, 0x02f502f5, 0x121202f5, 0xedee02f5, - 0x140502f5, 0xebfb02f5, 0x051402f5, 0x00001212, 0x08081212, 0xf7f81212, 0x0afd1212, 0xf5031212, - 0xfd0b1212, 0x02f51212, 0x12121212, 0xedee1212, 0x14051212, 0xebfb1212, 0x05141212, 0xffffedee, - 0x0807edee, 0xf7f7edee, 0x0afcedee, 0xf502edee, 0xfd0aedee, 0x02f4edee, 0x1211edee, 0xedededee, - 0x1404edee, 0xebfaedee, 0x0513edee, 0x00001405, 0x08081405, 0xf7f81405, 0x0afd1405, 0xf5031405, - 0xfd0b1405, 0x02f51405, 0x12121405, 0xedee1405, 0x14051405, 0xebfb1405, 0x05141405, 0xffffebfb, - 0x0807ebfb, 0xf7f7ebfb, 0x0afcebfb, 0xf502ebfb, 0xfd0aebfb, 0x02f4ebfb, 0x1211ebfb, 0xededebfb, - 0x1404ebfb, 0xebfaebfb, 0x0513ebfb, 0x00000514, 0x08080514, 0xf7f80514, 0x0afd0514, 0xf5030514, - 0xfd0b0514, 0x02f50514, 0x12120514, 0xedee0514, 0x14050514, 0xebfb0514, 0x05140514, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000909, 0xfffff6f7, 0x00000bfd, 0xfffff403, 0xfffffd0c, 0x000002f4, 0x00001414, - 0xffffebec, 0x00001706, 0xffffe8fa, 0x00000617, 0xfffff9e9, 0x000010ef, 0xffffef11, 0x00001af9, - 0xffffe507, 0xfffff91b, 0x000006e5, 0x00002713, 0xffffd8ed, 0x00001327, 0xffffecd9, 0x00002727, - 0xffffd8d9, 0x00002b03, 0xffffd4fd, 0x0000032b, 0xfffffcd5, 0x000023e8, 0xffffdc18, 0xffffe824, - 0x000017dc, 0x0000452a, 0xffffbad6, 0x00002a45, 0xffffd5bb, 0x000034f2, 0xffffcb0e, 0xfffff235, - 0x00000dcb, 0x00004713, 0xffffb8ed, 0x00001347, 0xffffecb9, 0x00004949, 0xffffb6b7, 0x00004ffe, - 0xffffb002, 0xfffffe50, 0x000001b0, 0x000033cc, 0xffffcc34, 0x000045d9, 0xffffba27, 0xffffd946, - 0x000026ba, 0x00007930, 0xffff86d0, 0x00003079, 0xffffcf87, 0x00007c54, 0xffff83ac, 0x0000547c, - 0xffffab84, 0x000063e5, 0xffff9c1b, 0xffffe564, 0x00001a9c, 0x000065af, 0xffff9a51, 0xffffaf66, - 0x0000509a, 0x00003737, 0xffffc8c9, 0x00006868, 0xffff9798, 0x00000000, 0x09090000, 0xf6f70000, - 0x0bfd0000, 0xf4030000, 0xfd0c0000, 0x02f40000, 0x14140000, 0xebec0000, 0x17060000, 0xe8fa0000, - 0x06170000, 0xf9e90000, 0x00000909, 0x09090909, 0xf6f70909, 0x0bfd0909, 0xf4030909, 0xfd0c0909, - 0x02f40909, 0x14140909, 0xebec0909, 0x17060909, 0xe8fa0909, 0x06170909, 0xf9e90909, 0xfffff6f7, - 0x0908f6f7, 0xf6f6f6f7, 0x0bfcf6f7, 0xf402f6f7, 0xfd0bf6f7, 0x02f3f6f7, 0x1413f6f7, 0xebebf6f7, - 0x1705f6f7, 0xe8f9f6f7, 0x0616f6f7, 0xf9e8f6f7, 0x00000bfd, 0x09090bfd, 0xf6f70bfd, 0x0bfd0bfd, - 0xf4030bfd, 0xfd0c0bfd, 0x02f40bfd, 0x14140bfd, 0xebec0bfd, 0x17060bfd, 0xe8fa0bfd, 0x06170bfd, - 0xf9e90bfd, 0xfffff403, 0x0908f403, 0xf6f6f403, 0x0bfcf403, 0xf402f403, 0xfd0bf403, 0x02f3f403, - 0x1413f403, 0xebebf403, 0x1705f403, 0xe8f9f403, 0x0616f403, 0xf9e8f403, 0xfffffd0c, 0x0908fd0c, - 0xf6f6fd0c, 0x0bfcfd0c, 0xf402fd0c, 0xfd0bfd0c, 0x02f3fd0c, 0x1413fd0c, 0xebebfd0c, 0x1705fd0c, - 0xe8f9fd0c, 0x0616fd0c, 0xf9e8fd0c, 0x000002f4, 0x090902f4, 0xf6f702f4, 0x0bfd02f4, 0xf40302f4, - 0xfd0c02f4, 0x02f402f4, 0x141402f4, 0xebec02f4, 0x170602f4, 0xe8fa02f4, 0x061702f4, 0xf9e902f4, - 0x00001414, 0x09091414, 0xf6f71414, 0x0bfd1414, 0xf4031414, 0xfd0c1414, 0x02f41414, 0x14141414, - 0xebec1414, 0x17061414, 0xe8fa1414, 0x06171414, 0xf9e91414, 0xffffebec, 0x0908ebec, 0xf6f6ebec, - 0x0bfcebec, 0xf402ebec, 0xfd0bebec, 0x02f3ebec, 0x1413ebec, 0xebebebec, 0x1705ebec, 0xe8f9ebec, - 0x0616ebec, 0xf9e8ebec, 0x00001706, 0x09091706, 0xf6f71706, 0x0bfd1706, 0xf4031706, 0xfd0c1706, - 0x02f41706, 0x14141706, 0xebec1706, 0x17061706, 0xe8fa1706, 0x06171706, 0xf9e91706, 0xffffe8fa, - 0x0908e8fa, 0xf6f6e8fa, 0x0bfce8fa, 0xf402e8fa, 0xfd0be8fa, 0x02f3e8fa, 0x1413e8fa, 0xebebe8fa, - 0x1705e8fa, 0xe8f9e8fa, 0x0616e8fa, 0xf9e8e8fa, 0x00000617, 0x09090617, 0xf6f70617, 0x0bfd0617, - 0xf4030617, 0xfd0c0617, 0x02f40617, 0x14140617, 0xebec0617, 0x17060617, 0xe8fa0617, 0x06170617, - 0xf9e90617, 0xfffff9e9, 0x0908f9e9, 0xf6f6f9e9, 0x0bfcf9e9, 0xf402f9e9, 0xfd0bf9e9, 0x02f3f9e9, - 0x1413f9e9, 0xebebf9e9, 0x1705f9e9, 0xe8f9f9e9, 0x0616f9e9, 0xf9e8f9e9, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, - 0xfffffbfc, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x000003fc, 0xfffffc04, 0x000005fe, - 0xfffffa02, 0xfffffe06, 0x000001fa, 0x00000804, 0xfffff7fc, 0x00000408, 0xfffffbf8, 0x00000808, - 0xfffff7f8, 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x000007fc, 0xfffff804, 0xfffffc08, - 0x000003f8, 0x00000e08, 0xfffff1f8, 0x0000080e, 0xfffff7f2, 0x00000bfe, 0xfffff402, 0xfffffe0c, - 0x000001f4, 0x00001004, 0xffffeffc, 0x00000410, 0xfffffbf0, 0x00001010, 0xffffeff0, 0x00001200, - 0xffffee00, 0x00000012, 0xffffffee, 0x00000bf4, 0xfffff40c, 0x00000ff8, 0xfffff008, 0xfffff810, - 0x000007f0, 0x00001a0a, 0xffffe5f6, 0x00000a1a, 0xfffff5e6, 0x00001c12, 0xffffe3ee, 0x0000121c, - 0xffffede4, 0x000015fa, 0xffffea06, 0xfffffa16, 0x000005ea, 0x00001c04, 0xffffe3fc, 0x0000041c, - 0xfffffbe4, 0x00001e1e, 0xffffe1e2, 0x00001ffe, 0xffffe002, 0xfffffe20, 0x000001e0, 0x000015ee, - 0xffffea12, 0xffffee16, 0x000011ea, 0x00001df2, 0xffffe20e, 0xfffff21e, 0x00000de2, 0x00002e16, - 0xffffd1ea, 0x0000162e, 0xffffe9d2, 0x00002e0c, 0xffffd1f4, 0x00000c2e, 0xfffff3d2, 0x00003022, - 0xffffcfde, 0x00002230, 0xffffddd0, 0x000027f6, 0xffffd80a, 0xfffff628, 0x000009d8, 0x00003204, - 0xffffcdfc, 0x00000432, 0xfffffbce, 0x00003636, 0xffffc9ca, 0x000021de, 0xffffde22, 0x000029e4, - 0xffffd61c, 0xffffe42a, 0x00001bd6, 0x00003bfa, 0xffffc406, 0xfffffa3c, 0x000005c4, 0x00004c1a, - 0xffffb3e6, 0x00001a4c, 0xffffe5b4, 0x00004c2a, 0xffffb3d6, 0x00002a4c, 0xffffd5b4, 0x000035e8, - 0xffffca18, 0xffffe836, 0x000017ca, 0x00004e0e, 0xffffb1f2, 0x00000e4e, 0xfffff1b2, 0x0000523e, - 0xffffadc2, 0x00003e52, 0xffffc1ae, 0x000049ec, 0xffffb614, 0xffffec4a, 0x000013b6, 0x00005802, - 0xffffa7fe, 0x00000258, 0xfffffda8, 0x00005c5c, 0xffffa3a4, 0x00003bcc, 0xffffc434, 0xffffcc3c, - 0x000033c4, 0x00007634, 0xffff89cc, 0x00003476, 0xffffcb8a, 0x000049d4, 0xffffb62c, 0xffffd44a, - 0x00002bb6, 0x0000764a, 0xffff89b6, 0x00004a76, 0xffffb58a, 0x00007620, 0xffff89e0, 0x00002076, - 0xffffdf8a, 0x000065f4, 0xffff9a0c, 0xfffff466, 0x00000b9a, 0x00005fd8, 0xffffa028, 0xffffd860, - 0x000027a0, 0x000075de, 0xffff8a22, 0xffffde76, 0x0000218a, 0x000057a8, 0xffffa858, 0x000067b2, - 0xffff984e, 0xffffb268, 0x00004d98, 0x00000c0c, 0xfffff3f4, 0x00001616, 0xffffe9ea, 0x00002a2a, - 0xffffd5d6, 0x00004848, 0xffffb7b8, 0x00000000, 0x02020000, 0xfdfe0000, 0x02000000, 0xfe000000, - 0x00020000, 0xfffe0000, 0x00000202, 0x02020202, 0xfdfe0202, 0x02000202, 0xfe000202, 0x00020202, - 0xfffe0202, 0xfffffdfe, 0x0201fdfe, 0xfdfdfdfe, 0x01fffdfe, 0xfdfffdfe, 0x0001fdfe, 0xfffdfdfe, - 0x00000200, 0x02020200, 0xfdfe0200, 0x02000200, 0xfe000200, 0x00020200, 0xfffe0200, 0xfffffe00, - 0x0201fe00, 0xfdfdfe00, 0x01fffe00, 0xfdfffe00, 0x0001fe00, 0xfffdfe00, 0x00000002, 0x02020002, - 0xfdfe0002, 0x02000002, 0xfe000002, 0x00020002, 0xfffe0002, 0xfffffffe, 0x0201fffe, 0xfdfdfffe, - 0x01fffffe, 0xfdfffffe, 0x0001fffe, 0xfffdfffe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000303, 0xfffffcfd, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, - 0xfffff9fa, 0x00000903, 0xfffff6fd, 0x00000309, 0xfffffcf7, 0x000008fd, 0xfffff703, 0xfffffd09, - 0x000002f7, 0x000005fa, 0xfffffa06, 0x00000c06, 0xfffff3fa, 0x0000060c, 0xfffff9f4, 0x00000c0c, - 0xfffff3f4, 0x00000f00, 0xfffff100, 0x0000000f, 0xfffffff1, 0x00000bf7, 0xfffff409, 0xfffff70c, - 0x000008f4, 0x0000180f, 0xffffe7f1, 0x00000f18, 0xfffff0e8, 0x000011fa, 0xffffee06, 0xfffffa12, - 0x000005ee, 0x00001806, 0xffffe7fa, 0x00000618, 0xfffff9e8, 0x00001818, 0xffffe7e8, 0x00001b00, - 0xffffe500, 0x0000001b, 0xffffffe5, 0x000011ee, 0xffffee12, 0x000017f4, 0xffffe80c, 0xfffff418, - 0x00000be8, 0x0000270f, 0xffffd8f1, 0x00000f27, 0xfffff0d9, 0x00002a1b, 0xffffd5e5, 0x00001b2a, - 0xffffe4d6, 0x000020f7, 0xffffdf09, 0xfffff721, 0x000008df, 0x00002a06, 0xffffd5fa, 0x0000062a, - 0xfffff9d6, 0x00002d2d, 0xffffd2d3, 0x000032fd, 0xffffcd03, 0xfffffd33, 0x000002cd, 0x000020e5, - 0xffffdf1b, 0xffffe521, 0x00001adf, 0x00002ceb, 0xffffd315, 0xffffeb2d, 0x000014d3, 0x00004521, - 0xffffbadf, 0x00002145, 0xffffdebb, 0x00004512, 0xffffbaee, 0x00001245, 0xffffedbb, 0x00004836, - 0xffffb7ca, 0x00003648, 0xffffc9b8, 0x00003eee, 0xffffc112, 0xffffee3f, 0x000011c1, 0x00004e06, - 0xffffb1fa, 0x0000064e, 0xfffff9b2, 0x00005151, 0xffffaeaf, 0x000032cd, 0xffffcd33, 0x00003ed6, - 0xffffc12a, 0xffffd63f, 0x000029c1, 0x000059f7, 0xffffa609, 0xfffff75a, 0x000008a6, 0x0000722a, - 0xffff8dd6, 0x00002a72, 0xffffd58e, 0x0000753f, 0xffff8ac1, 0x00003f75, 0xffffc08b, 0x000050dc, - 0xffffaf24, 0xffffdc51, 0x000023af, 0x00007815, 0xffff87eb, 0x00001578, 0xffffea88, 0x00007b60, - 0xffff84a0, 0x0000607b, 0xffff9f85, 0x00006ee2, 0xffff911e, 0xffffe26f, 0x00001d91, 0x00005cb2, - 0xffffa34e, 0xffffb25d, 0x00004da3, 0x000071bb, 0xffff8e45, 0xffffbb72, 0x0000448e, 0x00001212, - 0xffffedee, 0x00002121, 0xffffdedf, 0x00003f3f, 0xffffc0c1, 0x00006c6c, 0xffff9394, 0x00000000, - 0x03030000, 0xfcfd0000, 0x03000000, 0xfd000000, 0x00030000, 0xfffd0000, 0x06060000, 0xf9fa0000, - 0x00000303, 0x03030303, 0xfcfd0303, 0x03000303, 0xfd000303, 0x00030303, 0xfffd0303, 0x06060303, - 0xf9fa0303, 0xfffffcfd, 0x0302fcfd, 0xfcfcfcfd, 0x02fffcfd, 0xfcfffcfd, 0x0002fcfd, 0xfffcfcfd, - 0x0605fcfd, 0xf9f9fcfd, 0x00000300, 0x03030300, 0xfcfd0300, 0x03000300, 0xfd000300, 0x00030300, - 0xfffd0300, 0x06060300, 0xf9fa0300, 0xfffffd00, 0x0302fd00, 0xfcfcfd00, 0x02fffd00, 0xfcfffd00, - 0x0002fd00, 0xfffcfd00, 0x0605fd00, 0xf9f9fd00, 0x00000003, 0x03030003, 0xfcfd0003, 0x03000003, - 0xfd000003, 0x00030003, 0xfffd0003, 0x06060003, 0xf9fa0003, 0xfffffffd, 0x0302fffd, 0xfcfcfffd, - 0x02fffffd, 0xfcfffffd, 0x0002fffd, 0xfffcfffd, 0x0605fffd, 0xf9f9fffd, 0x00000606, 0x03030606, - 0xfcfd0606, 0x03000606, 0xfd000606, 0x00030606, 0xfffd0606, 0x06060606, 0xf9fa0606, 0xfffff9fa, - 0x0302f9fa, 0xfcfcf9fa, 0x02fff9fa, 0xfcfff9fa, 0x0002f9fa, 0xfffcf9fa, 0x0605f9fa, 0xf9f9f9fa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000404, 0xfffffbfc, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000804, - 0xfffff7fc, 0x00000408, 0xfffffbf8, 0x00000808, 0xfffff7f8, 0x000007f8, 0xfffff808, 0x00000bfc, - 0xfffff404, 0xfffffc0c, 0x000003f4, 0x00001008, 0xffffeff8, 0x00000810, 0xfffff7f0, 0x00001010, - 0xffffeff0, 0x00001400, 0xffffec00, 0x00000014, 0xffffffec, 0x00000ff4, 0xfffff00c, 0xfffff410, - 0x00000bf0, 0x000017fc, 0xffffe804, 0xfffffc18, 0x000003e8, 0x00002010, 0xffffdff0, 0x00001020, - 0xffffefe0, 0x00002008, 0xffffdff8, 0x00000820, 0xfffff7e0, 0x00002020, 0xffffdfe0, 0x00002400, - 0xffffdc00, 0x00000024, 0xffffffdc, 0x000017e8, 0xffffe818, 0x00001ff0, 0xffffe010, 0xfffff020, - 0x00000fe0, 0x00003414, 0xffffcbec, 0x00001434, 0xffffebcc, 0x00003824, 0xffffc7dc, 0x00002438, - 0xffffdbc8, 0x00002bf4, 0xffffd40c, 0xfffff42c, 0x00000bd4, 0x00003808, 0xffffc7f8, 0x00000838, - 0xfffff7c8, 0x00003c3c, 0xffffc3c4, 0x00003ffc, 0xffffc004, 0xfffffc40, 0x000003c0, 0x00002bdc, - 0xffffd424, 0xffffdc2c, 0x000023d4, 0x00003be4, 0xffffc41c, 0xffffe43c, 0x00001bc4, 0x00005c2c, - 0xffffa3d4, 0x00002c5c, 0xffffd3a4, 0x00005c18, 0xffffa3e8, 0x0000185c, 0xffffe7a4, 0x00006048, - 0xffff9fb8, 0x00004860, 0xffffb7a0, 0x000053ec, 0xffffac14, 0xffffec54, 0x000013ac, 0x00006408, - 0xffff9bf8, 0x00000864, 0xfffff79c, 0x00006c6c, 0xffff9394, 0x000043bc, 0xffffbc44, 0x000053c8, - 0xffffac38, 0xffffc854, 0x000037ac, 0x000077f4, 0xffff880c, 0xfffff478, 0x00000b88, 0x00006bd0, - 0xffff9430, 0xffffd06c, 0x00002f94, 0x00007b98, 0xffff8468, 0xffff987c, 0x00006784, 0x00001818, - 0xffffe7e8, 0x00002c2c, 0xffffd3d4, 0x00005454, 0xffffabac, 0x00000000, 0x04040000, 0xfbfc0000, - 0x04000000, 0xfc000000, 0x00040000, 0xfffc0000, 0x08040000, 0xf7fc0000, 0x04080000, 0x00000404, - 0x04040404, 0xfbfc0404, 0x04000404, 0xfc000404, 0x00040404, 0xfffc0404, 0x08040404, 0xf7fc0404, - 0x04080404, 0xfffffbfc, 0x0403fbfc, 0xfbfbfbfc, 0x03fffbfc, 0xfbfffbfc, 0x0003fbfc, 0xfffbfbfc, - 0x0803fbfc, 0xf7fbfbfc, 0x0407fbfc, 0x00000400, 0x04040400, 0xfbfc0400, 0x04000400, 0xfc000400, - 0x00040400, 0xfffc0400, 0x08040400, 0xf7fc0400, 0x04080400, 0xfffffc00, 0x0403fc00, 0xfbfbfc00, - 0x03fffc00, 0xfbfffc00, 0x0003fc00, 0xfffbfc00, 0x0803fc00, 0xf7fbfc00, 0x0407fc00, 0x00000004, - 0x04040004, 0xfbfc0004, 0x04000004, 0xfc000004, 0x00040004, 0xfffc0004, 0x08040004, 0xf7fc0004, - 0x04080004, 0xfffffffc, 0x0403fffc, 0xfbfbfffc, 0x03fffffc, 0xfbfffffc, 0x0003fffc, 0xfffbfffc, - 0x0803fffc, 0xf7fbfffc, 0x0407fffc, 0x00000804, 0x04040804, 0xfbfc0804, 0x04000804, 0xfc000804, - 0x00040804, 0xfffc0804, 0x08040804, 0xf7fc0804, 0x04080804, 0xfffff7fc, 0x0403f7fc, 0xfbfbf7fc, - 0x03fff7fc, 0xfbfff7fc, 0x0003f7fc, 0xfffbf7fc, 0x0803f7fc, 0xf7fbf7fc, 0x0407f7fc, 0x00000408, - 0x04040408, 0xfbfc0408, 0x04000408, 0xfc000408, 0x00040408, 0xfffc0408, 0x08040408, 0xf7fc0408, - 0x04080408, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000505, 0xfffffafb, 0x00000500, 0xfffffb00, 0x00000005, 0xfffffffb, 0x00000a0a, - 0xfffff5f6, 0x00000f05, 0xfffff0fb, 0x0000050f, 0xfffffaf1, 0x000009f6, 0xfffff60a, 0x00000efb, - 0xfffff105, 0xfffffb0f, 0x000004f1, 0x0000140a, 0xffffebf6, 0x00000a14, 0xfffff5ec, 0x00001414, - 0xffffebec, 0x00001900, 0xffffe700, 0x00000019, 0xffffffe7, 0x000013f1, 0xffffec0f, 0xfffff114, - 0x00000eec, 0x00002819, 0xffffd7e7, 0x00001928, 0xffffe6d8, 0x00001df6, 0xffffe20a, 0xfffff61e, - 0x000009e2, 0x0000280a, 0xffffd7f6, 0x00000a28, 0xfffff5d8, 0x00002828, 0xffffd7d8, 0x00002d00, - 0xffffd300, 0x0000002d, 0xffffffd3, 0x00001de2, 0xffffe21e, 0x000027ec, 0xffffd814, 0xffffec28, - 0x000013d8, 0x00004119, 0xffffbee7, 0x00001941, 0xffffe6bf, 0x0000462d, 0xffffb9d3, 0x00002d46, - 0xffffd2ba, 0x000036f1, 0xffffc90f, 0xfffff137, 0x00000ec9, 0x0000460a, 0xffffb9f6, 0x00000a46, - 0xfffff5ba, 0x00004b4b, 0xffffb4b5, 0x000054fb, 0xffffab05, 0xfffffb55, 0x000004ab, 0x000036d3, - 0xffffc92d, 0xffffd337, 0x00002cc9, 0x00004add, 0xffffb523, 0xffffdd4b, 0x000022b5, 0x00007337, - 0xffff8cc9, 0x00003773, 0xffffc88d, 0x0000731e, 0xffff8ce2, 0x00001e73, 0xffffe18d, 0x0000785a, - 0xffff87a6, 0x00005a78, 0xffffa588, 0x000068e2, 0xffff971e, 0xffffe269, 0x00001d97, 0x000054ab, - 0xffffab55, 0x000068ba, 0xffff9746, 0xffffba69, 0x00004597, 0x00001e1e, 0xffffe1e2, 0x00003c3c, - 0xffffc3c4, 0x00006969, 0xffff9697, 0x00000000, 0x05050000, 0xfafb0000, 0x05000000, 0xfb000000, - 0x00050000, 0xfffb0000, 0x0a0a0000, 0xf5f60000, 0x0f050000, 0xf0fb0000, 0x00000505, 0x05050505, - 0xfafb0505, 0x05000505, 0xfb000505, 0x00050505, 0xfffb0505, 0x0a0a0505, 0xf5f60505, 0x0f050505, - 0xf0fb0505, 0xfffffafb, 0x0504fafb, 0xfafafafb, 0x04fffafb, 0xfafffafb, 0x0004fafb, 0xfffafafb, - 0x0a09fafb, 0xf5f5fafb, 0x0f04fafb, 0xf0fafafb, 0x00000500, 0x05050500, 0xfafb0500, 0x05000500, - 0xfb000500, 0x00050500, 0xfffb0500, 0x0a0a0500, 0xf5f60500, 0x0f050500, 0xf0fb0500, 0xfffffb00, - 0x0504fb00, 0xfafafb00, 0x04fffb00, 0xfafffb00, 0x0004fb00, 0xfffafb00, 0x0a09fb00, 0xf5f5fb00, - 0x0f04fb00, 0xf0fafb00, 0x00000005, 0x05050005, 0xfafb0005, 0x05000005, 0xfb000005, 0x00050005, - 0xfffb0005, 0x0a0a0005, 0xf5f60005, 0x0f050005, 0xf0fb0005, 0xfffffffb, 0x0504fffb, 0xfafafffb, - 0x04fffffb, 0xfafffffb, 0x0004fffb, 0xfffafffb, 0x0a09fffb, 0xf5f5fffb, 0x0f04fffb, 0xf0fafffb, - 0x00000a0a, 0x05050a0a, 0xfafb0a0a, 0x05000a0a, 0xfb000a0a, 0x00050a0a, 0xfffb0a0a, 0x0a0a0a0a, - 0xf5f60a0a, 0x0f050a0a, 0xf0fb0a0a, 0xfffff5f6, 0x0504f5f6, 0xfafaf5f6, 0x04fff5f6, 0xfafff5f6, - 0x0004f5f6, 0xfffaf5f6, 0x0a09f5f6, 0xf5f5f5f6, 0x0f04f5f6, 0xf0faf5f6, 0x00000f05, 0x05050f05, - 0xfafb0f05, 0x05000f05, 0xfb000f05, 0x00050f05, 0xfffb0f05, 0x0a0a0f05, 0xf5f60f05, 0x0f050f05, - 0xf0fb0f05, 0xfffff0fb, 0x0504f0fb, 0xfafaf0fb, 0x04fff0fb, 0xfafff0fb, 0x0004f0fb, 0xfffaf0fb, - 0x0a09f0fb, 0xf5f5f0fb, 0x0f04f0fb, 0xf0faf0fb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000606, 0xfffff9fa, 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x00000c0c, - 0xfffff3f4, 0x00000c06, 0xfffff3fa, 0x0000060c, 0xfffff9f4, 0x00000bf4, 0xfffff40c, 0x000011fa, - 0xffffee06, 0xfffffa12, 0x000005ee, 0x0000180c, 0xffffe7f4, 0x00000c18, 0xfffff3e8, 0x00001818, - 0xffffe7e8, 0x00001e00, 0xffffe200, 0x0000001e, 0xffffffe2, 0x000017ee, 0xffffe812, 0xffffee18, - 0x000011e8, 0x0000301e, 0xffffcfe2, 0x00001e30, 0xffffe1d0, 0x000023fa, 0xffffdc06, 0xfffffa24, - 0x000005dc, 0x0000300c, 0xffffcff4, 0x00000c30, 0xfffff3d0, 0x00003030, 0xffffcfd0, 0x00003600, - 0xffffca00, 0x00000036, 0xffffffca, 0x000023dc, 0xffffdc24, 0x00002fe8, 0xffffd018, 0xffffe830, - 0x000017d0, 0x00004e1e, 0xffffb1e2, 0x00001e4e, 0xffffe1b2, 0x00005436, 0xffffabca, 0x00003654, - 0xffffc9ac, 0x000041ee, 0xffffbe12, 0xffffee42, 0x000011be, 0x0000540c, 0xffffabf4, 0x00000c54, - 0xfffff3ac, 0x00005a5a, 0xffffa5a6, 0x00005ffa, 0xffffa006, 0xfffffa60, 0x000005a0, 0x000041ca, - 0xffffbe36, 0xffffca42, 0x000035be, 0x000059d6, 0xffffa62a, 0xffffd65a, 0x000029a6, 0x00007de2, - 0xffff821e, 0xffffe27e, 0x00001d82, 0x0000659a, 0xffff9a66, 0x00007dac, 0xffff8254, 0xffffac7e, - 0x00005382, 0x00002424, 0xffffdbdc, 0x00004242, 0xffffbdbe, 0x00000000, 0x06060000, 0xf9fa0000, - 0x06000000, 0xfa000000, 0x00060000, 0xfffa0000, 0x0c0c0000, 0xf3f40000, 0x0c060000, 0xf3fa0000, - 0x060c0000, 0x00000606, 0x06060606, 0xf9fa0606, 0x06000606, 0xfa000606, 0x00060606, 0xfffa0606, - 0x0c0c0606, 0xf3f40606, 0x0c060606, 0xf3fa0606, 0x060c0606, 0xfffff9fa, 0x0605f9fa, 0xf9f9f9fa, - 0x05fff9fa, 0xf9fff9fa, 0x0005f9fa, 0xfff9f9fa, 0x0c0bf9fa, 0xf3f3f9fa, 0x0c05f9fa, 0xf3f9f9fa, - 0x060bf9fa, 0x00000600, 0x06060600, 0xf9fa0600, 0x06000600, 0xfa000600, 0x00060600, 0xfffa0600, - 0x0c0c0600, 0xf3f40600, 0x0c060600, 0xf3fa0600, 0x060c0600, 0xfffffa00, 0x0605fa00, 0xf9f9fa00, - 0x05fffa00, 0xf9fffa00, 0x0005fa00, 0xfff9fa00, 0x0c0bfa00, 0xf3f3fa00, 0x0c05fa00, 0xf3f9fa00, - 0x060bfa00, 0x00000006, 0x06060006, 0xf9fa0006, 0x06000006, 0xfa000006, 0x00060006, 0xfffa0006, - 0x0c0c0006, 0xf3f40006, 0x0c060006, 0xf3fa0006, 0x060c0006, 0xfffffffa, 0x0605fffa, 0xf9f9fffa, - 0x05fffffa, 0xf9fffffa, 0x0005fffa, 0xfff9fffa, 0x0c0bfffa, 0xf3f3fffa, 0x0c05fffa, 0xf3f9fffa, - 0x060bfffa, 0x00000c0c, 0x06060c0c, 0xf9fa0c0c, 0x06000c0c, 0xfa000c0c, 0x00060c0c, 0xfffa0c0c, - 0x0c0c0c0c, 0xf3f40c0c, 0x0c060c0c, 0xf3fa0c0c, 0x060c0c0c, 0xfffff3f4, 0x0605f3f4, 0xf9f9f3f4, - 0x05fff3f4, 0xf9fff3f4, 0x0005f3f4, 0xfff9f3f4, 0x0c0bf3f4, 0xf3f3f3f4, 0x0c05f3f4, 0xf3f9f3f4, - 0x060bf3f4, 0x00000c06, 0x06060c06, 0xf9fa0c06, 0x06000c06, 0xfa000c06, 0x00060c06, 0xfffa0c06, - 0x0c0c0c06, 0xf3f40c06, 0x0c060c06, 0xf3fa0c06, 0x060c0c06, 0xfffff3fa, 0x0605f3fa, 0xf9f9f3fa, - 0x05fff3fa, 0xf9fff3fa, 0x0005f3fa, 0xfff9f3fa, 0x0c0bf3fa, 0xf3f3f3fa, 0x0c05f3fa, 0xf3f9f3fa, - 0x060bf3fa, 0x0000060c, 0x0606060c, 0xf9fa060c, 0x0600060c, 0xfa00060c, 0x0006060c, 0xfffa060c, - 0x0c0c060c, 0xf3f4060c, 0x0c06060c, 0xf3fa060c, 0x060c060c, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000707, 0xfffff8f9, 0x00000700, 0xfffff900, 0x00000007, 0xfffffff9, 0x00000e0e, - 0xfffff1f2, 0x00001507, 0xffffeaf9, 0x00000715, 0xfffff8eb, 0x00000df2, 0xfffff20e, 0x000014f9, - 0xffffeb07, 0xfffff915, 0x000006eb, 0x00001c0e, 0xffffe3f2, 0x00000e1c, 0xfffff1e4, 0x00001c1c, - 0xffffe3e4, 0x00002300, 0xffffdd00, 0x00000023, 0xffffffdd, 0x00001beb, 0xffffe415, 0xffffeb1c, - 0x000014e4, 0x00003823, 0xffffc7dd, 0x00002338, 0xffffdcc8, 0x000029f2, 0xffffd60e, 0xfffff22a, - 0x00000dd6, 0x0000380e, 0xffffc7f2, 0x00000e38, 0xfffff1c8, 0x00003838, 0xffffc7c8, 0x00003f00, - 0xffffc100, 0x0000003f, 0xffffffc1, 0x000029d6, 0xffffd62a, 0x000037e4, 0xffffc81c, 0xffffe438, - 0x00001bc8, 0x00005b23, 0xffffa4dd, 0x0000235b, 0xffffdca5, 0x0000623f, 0xffff9dc1, 0x00003f62, - 0xffffc09e, 0x00004ceb, 0xffffb315, 0xffffeb4d, 0x000014b3, 0x0000620e, 0xffff9df2, 0x00000e62, - 0xfffff19e, 0x00006969, 0xffff9697, 0x000076f9, 0xffff8907, 0xfffff977, 0x00000689, 0x00004cc1, - 0xffffb33f, 0xffffc14d, 0x00003eb3, 0x000068cf, 0xffff9731, 0xffffcf69, 0x00003097, 0x00007689, - 0xffff8977, 0x00002a2a, 0xffffd5d6, 0x00004d4d, 0xffffb2b3, 0x00000000, 0x07070000, 0xf8f90000, - 0x07000000, 0xf9000000, 0x00070000, 0xfff90000, 0x0e0e0000, 0xf1f20000, 0x15070000, 0xeaf90000, - 0x07150000, 0x00000707, 0x07070707, 0xf8f90707, 0x07000707, 0xf9000707, 0x00070707, 0xfff90707, - 0x0e0e0707, 0xf1f20707, 0x15070707, 0xeaf90707, 0x07150707, 0xfffff8f9, 0x0706f8f9, 0xf8f8f8f9, - 0x06fff8f9, 0xf8fff8f9, 0x0006f8f9, 0xfff8f8f9, 0x0e0df8f9, 0xf1f1f8f9, 0x1506f8f9, 0xeaf8f8f9, - 0x0714f8f9, 0x00000700, 0x07070700, 0xf8f90700, 0x07000700, 0xf9000700, 0x00070700, 0xfff90700, - 0x0e0e0700, 0xf1f20700, 0x15070700, 0xeaf90700, 0x07150700, 0xfffff900, 0x0706f900, 0xf8f8f900, - 0x06fff900, 0xf8fff900, 0x0006f900, 0xfff8f900, 0x0e0df900, 0xf1f1f900, 0x1506f900, 0xeaf8f900, - 0x0714f900, 0x00000007, 0x07070007, 0xf8f90007, 0x07000007, 0xf9000007, 0x00070007, 0xfff90007, - 0x0e0e0007, 0xf1f20007, 0x15070007, 0xeaf90007, 0x07150007, 0xfffffff9, 0x0706fff9, 0xf8f8fff9, - 0x06fffff9, 0xf8fffff9, 0x0006fff9, 0xfff8fff9, 0x0e0dfff9, 0xf1f1fff9, 0x1506fff9, 0xeaf8fff9, - 0x0714fff9, 0x00000e0e, 0x07070e0e, 0xf8f90e0e, 0x07000e0e, 0xf9000e0e, 0x00070e0e, 0xfff90e0e, - 0x0e0e0e0e, 0xf1f20e0e, 0x15070e0e, 0xeaf90e0e, 0x07150e0e, 0xfffff1f2, 0x0706f1f2, 0xf8f8f1f2, - 0x06fff1f2, 0xf8fff1f2, 0x0006f1f2, 0xfff8f1f2, 0x0e0df1f2, 0xf1f1f1f2, 0x1506f1f2, 0xeaf8f1f2, - 0x0714f1f2, 0x00001507, 0x07071507, 0xf8f91507, 0x07001507, 0xf9001507, 0x00071507, 0xfff91507, - 0x0e0e1507, 0xf1f21507, 0x15071507, 0xeaf91507, 0x07151507, 0xffffeaf9, 0x0706eaf9, 0xf8f8eaf9, - 0x06ffeaf9, 0xf8ffeaf9, 0x0006eaf9, 0xfff8eaf9, 0x0e0deaf9, 0xf1f1eaf9, 0x1506eaf9, 0xeaf8eaf9, - 0x0714eaf9, 0x00000715, 0x07070715, 0xf8f90715, 0x07000715, 0xf9000715, 0x00070715, 0xfff90715, - 0x0e0e0715, 0xf1f20715, 0x15070715, 0xeaf90715, 0x07150715, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000808, 0xfffff7f8, 0x00000800, 0xfffff800, 0x00000008, 0xfffffff8, 0x00001010, - 0xffffeff0, 0x00001008, 0xffffeff8, 0x00000810, 0xfffff7f0, 0x00000ff0, 0xfffff010, 0x000017f8, - 0xffffe808, 0xfffff818, 0x000007e8, 0x00002010, 0xffffdff0, 0x00001020, 0xffffefe0, 0x00002020, - 0xffffdfe0, 0x00002800, 0xffffd800, 0x00000028, 0xffffffd8, 0x00001fe8, 0xffffe018, 0xffffe820, - 0x000017e0, 0x00004028, 0xffffbfd8, 0x00002840, 0xffffd7c0, 0x00002ff0, 0xffffd010, 0xfffff030, - 0x00000fd0, 0x00004010, 0xffffbff0, 0x00001040, 0xffffefc0, 0x00004040, 0xffffbfc0, 0x00004800, - 0xffffb800, 0x00000048, 0xffffffb8, 0x00002fd0, 0xffffd030, 0x00003fe0, 0xffffc020, 0xffffe040, - 0x00001fc0, 0x00006828, 0xffff97d8, 0x00002868, 0xffffd798, 0x00007048, 0xffff8fb8, 0x00004870, - 0xffffb790, 0x000057e8, 0xffffa818, 0xffffe858, 0x000017a8, 0x00007010, 0xffff8ff0, 0x00001070, - 0xffffef90, 0x00007878, 0xffff8788, 0x000057b8, 0xffffa848, 0xffffb858, 0x000047a8, 0x000077c8, - 0xffff8838, 0xffffc878, 0x00003788, 0x00003030, 0xffffcfd0, 0x00005858, 0xffffa7a8, 0x00000000, - 0x08080000, 0xf7f80000, 0x08000000, 0xf8000000, 0x00080000, 0xfff80000, 0x10100000, 0xeff00000, - 0x10080000, 0xeff80000, 0x08100000, 0x00000808, 0x08080808, 0xf7f80808, 0x08000808, 0xf8000808, - 0x00080808, 0xfff80808, 0x10100808, 0xeff00808, 0x10080808, 0xeff80808, 0x08100808, 0xfffff7f8, - 0x0807f7f8, 0xf7f7f7f8, 0x07fff7f8, 0xf7fff7f8, 0x0007f7f8, 0xfff7f7f8, 0x100ff7f8, 0xefeff7f8, - 0x1007f7f8, 0xeff7f7f8, 0x080ff7f8, 0x00000800, 0x08080800, 0xf7f80800, 0x08000800, 0xf8000800, - 0x00080800, 0xfff80800, 0x10100800, 0xeff00800, 0x10080800, 0xeff80800, 0x08100800, 0xfffff800, - 0x0807f800, 0xf7f7f800, 0x07fff800, 0xf7fff800, 0x0007f800, 0xfff7f800, 0x100ff800, 0xefeff800, - 0x1007f800, 0xeff7f800, 0x080ff800, 0x00000008, 0x08080008, 0xf7f80008, 0x08000008, 0xf8000008, - 0x00080008, 0xfff80008, 0x10100008, 0xeff00008, 0x10080008, 0xeff80008, 0x08100008, 0xfffffff8, - 0x0807fff8, 0xf7f7fff8, 0x07fffff8, 0xf7fffff8, 0x0007fff8, 0xfff7fff8, 0x100ffff8, 0xefeffff8, - 0x1007fff8, 0xeff7fff8, 0x080ffff8, 0x00001010, 0x08081010, 0xf7f81010, 0x08001010, 0xf8001010, - 0x00081010, 0xfff81010, 0x10101010, 0xeff01010, 0x10081010, 0xeff81010, 0x08101010, 0xffffeff0, - 0x0807eff0, 0xf7f7eff0, 0x07ffeff0, 0xf7ffeff0, 0x0007eff0, 0xfff7eff0, 0x100feff0, 0xefefeff0, - 0x1007eff0, 0xeff7eff0, 0x080feff0, 0x00001008, 0x08081008, 0xf7f81008, 0x08001008, 0xf8001008, - 0x00081008, 0xfff81008, 0x10101008, 0xeff01008, 0x10081008, 0xeff81008, 0x08101008, 0xffffeff8, - 0x0807eff8, 0xf7f7eff8, 0x07ffeff8, 0xf7ffeff8, 0x0007eff8, 0xfff7eff8, 0x100feff8, 0xefefeff8, - 0x1007eff8, 0xeff7eff8, 0x080feff8, 0x00000810, 0x08080810, 0xf7f80810, 0x08000810, 0xf8000810, - 0x00080810, 0xfff80810, 0x10100810, 0xeff00810, 0x10080810, 0xeff80810, 0x08100810, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000909, 0xfffff6f7, 0x00000900, 0xfffff700, 0x00000009, 0xfffffff7, 0x00001212, - 0xffffedee, 0x00001b09, 0xffffe4f7, 0x0000091b, 0xfffff6e5, 0x000011ee, 0xffffee12, 0x00001af7, - 0xffffe509, 0xfffff71b, 0x000008e5, 0x00002412, 0xffffdbee, 0x00001224, 0xffffeddc, 0x00002424, - 0xffffdbdc, 0x00002d00, 0xffffd300, 0x0000002d, 0xffffffd3, 0x000023e5, 0xffffdc1b, 0xffffe524, - 0x00001adc, 0x0000482d, 0xffffb7d3, 0x00002d48, 0xffffd2b8, 0x000035ee, 0xffffca12, 0xffffee36, - 0x000011ca, 0x00004812, 0xffffb7ee, 0x00001248, 0xffffedb8, 0x00004848, 0xffffb7b8, 0x00005100, - 0xffffaf00, 0x00000051, 0xffffffaf, 0x000035ca, 0xffffca36, 0x000047dc, 0xffffb824, 0xffffdc48, - 0x000023b8, 0x0000752d, 0xffff8ad3, 0x00002d75, 0xffffd28b, 0x00007e51, 0xffff81af, 0x0000517e, - 0xffffae82, 0x000062e5, 0xffff9d1b, 0xffffe563, 0x00001a9d, 0x000062af, 0xffff9d51, 0xffffaf63, - 0x0000509d, 0x00003636, 0xffffc9ca, 0x00006c6c, 0xffff9394, 0x00000000, 0x09090000, 0xf6f70000, - 0x09000000, 0xf7000000, 0x00090000, 0xfff70000, 0x12120000, 0xedee0000, 0x1b090000, 0xe4f70000, - 0x091b0000, 0xf6e50000, 0x00000909, 0x09090909, 0xf6f70909, 0x09000909, 0xf7000909, 0x00090909, - 0xfff70909, 0x12120909, 0xedee0909, 0x1b090909, 0xe4f70909, 0x091b0909, 0xf6e50909, 0xfffff6f7, - 0x0908f6f7, 0xf6f6f6f7, 0x08fff6f7, 0xf6fff6f7, 0x0008f6f7, 0xfff6f6f7, 0x1211f6f7, 0xededf6f7, - 0x1b08f6f7, 0xe4f6f6f7, 0x091af6f7, 0xf6e4f6f7, 0x00000900, 0x09090900, 0xf6f70900, 0x09000900, - 0xf7000900, 0x00090900, 0xfff70900, 0x12120900, 0xedee0900, 0x1b090900, 0xe4f70900, 0x091b0900, - 0xf6e50900, 0xfffff700, 0x0908f700, 0xf6f6f700, 0x08fff700, 0xf6fff700, 0x0008f700, 0xfff6f700, - 0x1211f700, 0xededf700, 0x1b08f700, 0xe4f6f700, 0x091af700, 0xf6e4f700, 0x00000009, 0x09090009, - 0xf6f70009, 0x09000009, 0xf7000009, 0x00090009, 0xfff70009, 0x12120009, 0xedee0009, 0x1b090009, - 0xe4f70009, 0x091b0009, 0xf6e50009, 0xfffffff7, 0x0908fff7, 0xf6f6fff7, 0x08fffff7, 0xf6fffff7, - 0x0008fff7, 0xfff6fff7, 0x1211fff7, 0xededfff7, 0x1b08fff7, 0xe4f6fff7, 0x091afff7, 0xf6e4fff7, - 0x00001212, 0x09091212, 0xf6f71212, 0x09001212, 0xf7001212, 0x00091212, 0xfff71212, 0x12121212, - 0xedee1212, 0x1b091212, 0xe4f71212, 0x091b1212, 0xf6e51212, 0xffffedee, 0x0908edee, 0xf6f6edee, - 0x08ffedee, 0xf6ffedee, 0x0008edee, 0xfff6edee, 0x1211edee, 0xedededee, 0x1b08edee, 0xe4f6edee, - 0x091aedee, 0xf6e4edee, 0x00001b09, 0x09091b09, 0xf6f71b09, 0x09001b09, 0xf7001b09, 0x00091b09, - 0xfff71b09, 0x12121b09, 0xedee1b09, 0x1b091b09, 0xe4f71b09, 0x091b1b09, 0xf6e51b09, 0xffffe4f7, - 0x0908e4f7, 0xf6f6e4f7, 0x08ffe4f7, 0xf6ffe4f7, 0x0008e4f7, 0xfff6e4f7, 0x1211e4f7, 0xedede4f7, - 0x1b08e4f7, 0xe4f6e4f7, 0x091ae4f7, 0xf6e4e4f7, 0x0000091b, 0x0909091b, 0xf6f7091b, 0x0900091b, - 0xf700091b, 0x0009091b, 0xfff7091b, 0x1212091b, 0xedee091b, 0x1b09091b, 0xe4f7091b, 0x091b091b, - 0xf6e5091b, 0xfffff6e5, 0x0908f6e5, 0xf6f6f6e5, 0x08fff6e5, 0xf6fff6e5, 0x0008f6e5, 0xfff6f6e5, - 0x1211f6e5, 0xededf6e5, 0x1b08f6e5, 0xe4f6f6e5, 0x091af6e5, 0xf6e4f6e5, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, - 0xfffff9fa, 0x00000700, 0xfffff900, 0x00000007, 0xfffffff9, 0x000004fb, 0xfffffb05, 0xfffffb05, - 0x000004fb, 0x00000b06, 0xfffff4fa, 0x0000060b, 0xfffff9f5, 0x00000800, 0xfffff800, 0x00000008, - 0xfffffff8, 0x00000b0b, 0xfffff4f5, 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x0000110c, - 0xffffeef4, 0x00000c11, 0xfffff3ef, 0x00001111, 0xffffeeef, 0x00001206, 0xffffedfa, 0x00000612, - 0xfffff9ee, 0x00000af8, 0xfffff508, 0xfffff80b, 0x000007f5, 0x00000f00, 0xfffff100, 0x0000000f, - 0xfffffff1, 0x00001400, 0xffffec00, 0x00000014, 0xffffffec, 0x00001912, 0xffffe6ee, 0x00001219, - 0xffffede7, 0x0000190b, 0xffffe6f5, 0x00000b19, 0xfffff4e7, 0x00001919, 0xffffe6e7, 0x00000df2, - 0xfffff20e, 0xfffff20e, 0x00000df2, 0x00001a00, 0xffffe600, 0x0000001a, 0xffffffe6, 0x000011f5, - 0xffffee0b, 0xfffff512, 0x00000aee, 0x000015f9, 0xffffea07, 0xfffff916, 0x000006ea, 0x0000221a, - 0xffffdde6, 0x00001a22, 0xffffe5de, 0x00002212, 0xffffddee, 0x00001222, 0xffffedde, 0x00002222, - 0xffffddde, 0x0000230b, 0xffffdcf5, 0x00000b23, 0xfffff4dd, 0x00001d00, 0xffffe300, 0x0000001d, - 0xffffffe3, 0x000015ed, 0xffffea13, 0xffffed16, 0x000012ea, 0x000019f1, 0xffffe60f, 0xfffff11a, - 0x00000ee6, 0x00002500, 0xffffdb00, 0x00000025, 0xffffffdb, 0x00002c1b, 0xffffd3e5, 0x00001b2c, - 0xffffe4d4, 0x00002c24, 0xffffd3dc, 0x0000242c, 0xffffdbd4, 0x00002c12, 0xffffd3ee, 0x0000122c, - 0xffffedd4, 0x000020f6, 0xffffdf0a, 0xfffff621, 0x000009df, 0x00002d2d, 0xffffd2d3, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, - 0xfffff9fa, 0x00000700, 0xfffff900, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020300, 0x0201fd00, - 0x02020003, 0x0201fffd, 0x02020606, 0x0201f9fa, 0x02020700, 0x0201f900, 0xfdfe0000, 0xfdfe0202, - 0xfdfdfdfe, 0xfdfe0300, 0xfdfdfd00, 0xfdfe0003, 0xfdfdfffd, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0700, - 0xfdfdf900, 0x03000000, 0x03000202, 0x02fffdfe, 0x03000300, 0x02fffd00, 0x03000003, 0x02fffffd, - 0x03000606, 0x02fff9fa, 0x03000700, 0x02fff900, 0xfd000000, 0xfd000202, 0xfcfffdfe, 0xfd000300, - 0xfcfffd00, 0xfd000003, 0xfcfffffd, 0xfd000606, 0xfcfff9fa, 0xfd000700, 0xfcfff900, 0x00030000, - 0x00030202, 0x0002fdfe, 0x00030300, 0x0002fd00, 0x00030003, 0x0002fffd, 0x00030606, 0x0002f9fa, - 0x00030700, 0x0002f900, 0xfffd0000, 0xfffd0202, 0xfffcfdfe, 0xfffd0300, 0xfffcfd00, 0xfffd0003, - 0xfffcfffd, 0xfffd0606, 0xfffcf9fa, 0xfffd0700, 0xfffcf900, 0x06060000, 0x06060202, 0x0605fdfe, - 0x06060300, 0x0605fd00, 0x06060003, 0x0605fffd, 0x06060606, 0x0605f9fa, 0x06060700, 0x0605f900, - 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0300, 0xf9f9fd00, 0xf9fa0003, 0xf9f9fffd, 0xf9fa0606, - 0xf9f9f9fa, 0xf9fa0700, 0xf9f9f900, 0x07000000, 0x07000202, 0x06fffdfe, 0x07000300, 0x06fffd00, - 0x07000003, 0x06fffffd, 0x07000606, 0x06fff9fa, 0x07000700, 0x06fff900, 0xf9000000, 0xf9000202, - 0xf8fffdfe, 0xf9000300, 0xf8fffd00, 0xf9000003, 0xf8fffffd, 0xf9000606, 0xf8fff9fa, 0xf9000700, - 0xf8fff900, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000202, 0xfffffdfe, 0x00000606, - 0xfffff9fa, 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x000003fc, 0xfffffc04, 0xfffffa0a, - 0x000005f6, 0xfffff400, 0x00000c00, 0xfffff3fa, 0xfffff406, 0x00000bfa, 0x00000c06, 0xfffffff2, - 0x0000000e, 0x00000c0c, 0xfffff3f4, 0xffffee00, 0x00001200, 0xfffff40e, 0x00000bf2, 0xfffff9ee, - 0xfffffa12, 0x000005ee, 0x00000612, 0xffffedf6, 0xffffee0a, 0x000011f6, 0x0000120a, 0xffffffea, - 0x00000016, 0xffffe800, 0x00001800, 0xfffff3ea, 0xfffff416, 0x00000bea, 0x00000c16, 0xffffe7f8, - 0xffffe808, 0x000017f8, 0x00001808, 0xfffff9e6, 0xfffffa1a, 0x000005e6, 0x0000061a, 0xffffffe4, - 0x0000001c, 0x00001414, 0xffffebec, 0xffffe5f2, 0x00001a0e, 0xfffff3e2, 0x00000c1e, 0xffffdff6, - 0x0000200a, 0xffffdfee, 0x00002012, 0xffffe5e6, 0x00001a1a, 0xffffebde, 0x00001422, 0xfffff3da, - 0x00000c26, 0xffffdfe0, 0x00002020, 0x00002020, 0xffffd7ea, 0xffffddde, 0x00002222, 0x00000000, - 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, - 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x02000000, 0x02000200, 0x01fffe00, 0x02000002, - 0x01fffffe, 0x02000202, 0x01fffdfe, 0x02000606, 0x01fff9fa, 0x02000600, 0x01fffa00, 0x02000006, - 0x01fffffa, 0xfe000000, 0xfe000200, 0xfdfffe00, 0xfe000002, 0xfdfffffe, 0xfe000202, 0xfdfffdfe, - 0xfe000606, 0xfdfff9fa, 0xfe000600, 0xfdfffa00, 0xfe000006, 0xfdfffffa, 0x00020000, 0x00020200, - 0x0001fe00, 0x00020002, 0x0001fffe, 0x00020202, 0x0001fdfe, 0x00020606, 0x0001f9fa, 0x00020600, - 0x0001fa00, 0x00020006, 0x0001fffa, 0xfffe0000, 0xfffe0200, 0xfffdfe00, 0xfffe0002, 0xfffdfffe, - 0xfffe0202, 0xfffdfdfe, 0xfffe0606, 0xfffdf9fa, 0xfffe0600, 0xfffdfa00, 0xfffe0006, 0xfffdfffa, - 0x02020000, 0x02020200, 0x0201fe00, 0x02020002, 0x0201fffe, 0x02020202, 0x0201fdfe, 0x02020606, - 0x0201f9fa, 0x02020600, 0x0201fa00, 0x02020006, 0x0201fffa, 0xfdfe0000, 0xfdfe0200, 0xfdfdfe00, - 0xfdfe0002, 0xfdfdfffe, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0600, 0xfdfdfa00, - 0xfdfe0006, 0xfdfdfffa, 0x06060000, 0x06060200, 0x0605fe00, 0x06060002, 0x0605fffe, 0x06060202, - 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060600, 0x0605fa00, 0x06060006, 0x0605fffa, 0xf9fa0000, - 0xf9fa0200, 0xf9f9fe00, 0xf9fa0002, 0xf9f9fffe, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, - 0xf9fa0600, 0xf9f9fa00, 0xf9fa0006, 0xf9f9fffa, 0x06000000, 0x06000200, 0x05fffe00, 0x06000002, - 0x05fffffe, 0x06000202, 0x05fffdfe, 0x06000606, 0x05fff9fa, 0x06000600, 0x05fffa00, 0x06000006, - 0x05fffffa, 0xfa000000, 0xfa000200, 0xf9fffe00, 0xfa000002, 0xf9fffffe, 0xfa000202, 0xf9fffdfe, - 0xfa000606, 0xf9fff9fa, 0xfa000600, 0xf9fffa00, 0xfa000006, 0xf9fffffa, 0x00060000, 0x00060200, - 0x0005fe00, 0x00060002, 0x0005fffe, 0x00060202, 0x0005fdfe, 0x00060606, 0x0005f9fa, 0x00060600, - 0x0005fa00, 0x00060006, 0x0005fffa, 0xfffa0000, 0xfffa0200, 0xfff9fe00, 0xfffa0002, 0xfff9fffe, - 0xfffa0202, 0xfff9fdfe, 0xfffa0606, 0xfff9f9fa, 0xfffa0600, 0xfff9fa00, 0xfffa0006, 0xfff9fffa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, 0xfffffbfc, 0x00000a0a, - 0xfffff5f6, 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x000005fa, 0xfffffa06, 0xfffff80e, - 0x000007f2, 0xffffffee, 0x00000012, 0xfffff00a, 0x00000ff6, 0xffffe800, 0x00001800, 0xfffff7e8, - 0xfffff818, 0x000007e8, 0x00000818, 0x00001212, 0xffffedee, 0xfffff014, 0x00000fec, 0xffffe5f2, - 0xffffe60e, 0x000019f2, 0x00001a0e, 0xffffffe2, 0x0000001e, 0xffffde00, 0x00002200, 0xfffff7de, - 0xfffff822, 0x000007de, 0x00000822, 0xffffede2, 0xffffee1e, 0x000011e2, 0x0000121e, 0xffffddf6, - 0xffffde0a, 0x000021f6, 0x0000220a, 0xffffddec, 0x00002214, 0xffffffd8, 0x00000028, 0x00001e1e, - 0xffffe1e2, 0xffffedd8, 0x00001228, 0xffffd400, 0x00002c00, 0xffffd3f0, 0x00002c10, 0xffffdbdc, - 0xffffdbdc, 0x00002424, 0xffffd3e6, 0x00002c1a, 0xffffe5d2, 0x00001a2e, 0xffffedcc, 0x00001234, - 0xffffc9ec, 0xffffd3d4, 0x00002c2c, 0xffffc9e0, 0xffffd1d2, 0xffffd1d2, 0x00002e2e, 0x00000000, - 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, 0xfffffbfc, 0x00000a0a, 0xfffff5f6, - 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x02000000, 0x02000200, 0x01fffe00, 0x02000002, - 0x01fffffe, 0x02000404, 0x01fffbfc, 0x02000a0a, 0x01fff5f6, 0x02000a00, 0x01fff600, 0x0200000a, - 0x01fffff6, 0xfe000000, 0xfe000200, 0xfdfffe00, 0xfe000002, 0xfdfffffe, 0xfe000404, 0xfdfffbfc, - 0xfe000a0a, 0xfdfff5f6, 0xfe000a00, 0xfdfff600, 0xfe00000a, 0xfdfffff6, 0x00020000, 0x00020200, - 0x0001fe00, 0x00020002, 0x0001fffe, 0x00020404, 0x0001fbfc, 0x00020a0a, 0x0001f5f6, 0x00020a00, - 0x0001f600, 0x0002000a, 0x0001fff6, 0xfffe0000, 0xfffe0200, 0xfffdfe00, 0xfffe0002, 0xfffdfffe, - 0xfffe0404, 0xfffdfbfc, 0xfffe0a0a, 0xfffdf5f6, 0xfffe0a00, 0xfffdf600, 0xfffe000a, 0xfffdfff6, - 0x04040000, 0x04040200, 0x0403fe00, 0x04040002, 0x0403fffe, 0x04040404, 0x0403fbfc, 0x04040a0a, - 0x0403f5f6, 0x04040a00, 0x0403f600, 0x0404000a, 0x0403fff6, 0xfbfc0000, 0xfbfc0200, 0xfbfbfe00, - 0xfbfc0002, 0xfbfbfffe, 0xfbfc0404, 0xfbfbfbfc, 0xfbfc0a0a, 0xfbfbf5f6, 0xfbfc0a00, 0xfbfbf600, - 0xfbfc000a, 0xfbfbfff6, 0x0a0a0000, 0x0a0a0200, 0x0a09fe00, 0x0a0a0002, 0x0a09fffe, 0x0a0a0404, - 0x0a09fbfc, 0x0a0a0a0a, 0x0a09f5f6, 0x0a0a0a00, 0x0a09f600, 0x0a0a000a, 0x0a09fff6, 0xf5f60000, - 0xf5f60200, 0xf5f5fe00, 0xf5f60002, 0xf5f5fffe, 0xf5f60404, 0xf5f5fbfc, 0xf5f60a0a, 0xf5f5f5f6, - 0xf5f60a00, 0xf5f5f600, 0xf5f6000a, 0xf5f5fff6, 0x0a000000, 0x0a000200, 0x09fffe00, 0x0a000002, - 0x09fffffe, 0x0a000404, 0x09fffbfc, 0x0a000a0a, 0x09fff5f6, 0x0a000a00, 0x09fff600, 0x0a00000a, - 0x09fffff6, 0xf6000000, 0xf6000200, 0xf5fffe00, 0xf6000002, 0xf5fffffe, 0xf6000404, 0xf5fffbfc, - 0xf6000a0a, 0xf5fff5f6, 0xf6000a00, 0xf5fff600, 0xf600000a, 0xf5fffff6, 0x000a0000, 0x000a0200, - 0x0009fe00, 0x000a0002, 0x0009fffe, 0x000a0404, 0x0009fbfc, 0x000a0a0a, 0x0009f5f6, 0x000a0a00, - 0x0009f600, 0x000a000a, 0x0009fff6, 0xfff60000, 0xfff60200, 0xfff5fe00, 0xfff60002, 0xfff5fffe, - 0xfff60404, 0xfff5fbfc, 0xfff60a0a, 0xfff5f5f6, 0xfff60a00, 0xfff5f600, 0xfff6000a, 0xfff5fff6, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000404, 0xfffffbfc, 0x00000c0c, - 0xfffff3f4, 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x000007f8, 0xfffff808, 0xfffff008, - 0x00000ff8, 0xffffe800, 0x00001800, 0xfffff7e8, 0xfffff818, 0x000007e8, 0x00000818, 0xfffff014, - 0x00000fec, 0xffffffe4, 0x0000001c, 0xffffe7f0, 0xffffe810, 0x000017f0, 0x00001810, 0xffffe000, - 0x00002000, 0xffffefe4, 0xfffff01c, 0x00000fe4, 0x0000101c, 0xffffdff8, 0xffffe008, 0xfffff7e0, - 0xfffff820, 0x000007e0, 0x00000820, 0x00001ff8, 0x00002008, 0x00001818, 0xffffe7e8, 0xffffe818, - 0x000017e8, 0xffffdfec, 0x00002014, 0xffffffd8, 0x00000028, 0xffffefd8, 0x00001028, 0xffffd400, - 0xffffd400, 0xffffffd4, 0x0000002c, 0x00002c00, 0x00002c00, 0xffffdfe0, 0x00002020, 0xffffd3f0, - 0x00002c10, 0xffffd3e8, 0xffffe7d4, 0x0000182c, 0x00002c18, 0xffffefd0, 0x00001030, 0xffffdbdc, - 0xffffdbdc, 0x00002424, 0x00002424, 0xffffcbec, 0x00002828, 0xffffd7d8, 0xffffcbe0, 0x00000000, - 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000404, 0xfffffbfc, 0x00000c0c, 0xfffff3f4, - 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x04000000, 0x04000400, 0x03fffc00, 0x04000004, - 0x03fffffc, 0x04000404, 0x03fffbfc, 0x04000c0c, 0x03fff3f4, 0x04000c00, 0x03fff400, 0x0400000c, - 0x03fffff4, 0xfc000000, 0xfc000400, 0xfbfffc00, 0xfc000004, 0xfbfffffc, 0xfc000404, 0xfbfffbfc, - 0xfc000c0c, 0xfbfff3f4, 0xfc000c00, 0xfbfff400, 0xfc00000c, 0xfbfffff4, 0x00040000, 0x00040400, - 0x0003fc00, 0x00040004, 0x0003fffc, 0x00040404, 0x0003fbfc, 0x00040c0c, 0x0003f3f4, 0x00040c00, - 0x0003f400, 0x0004000c, 0x0003fff4, 0xfffc0000, 0xfffc0400, 0xfffbfc00, 0xfffc0004, 0xfffbfffc, - 0xfffc0404, 0xfffbfbfc, 0xfffc0c0c, 0xfffbf3f4, 0xfffc0c00, 0xfffbf400, 0xfffc000c, 0xfffbfff4, - 0x04040000, 0x04040400, 0x0403fc00, 0x04040004, 0x0403fffc, 0x04040404, 0x0403fbfc, 0x04040c0c, - 0x0403f3f4, 0x04040c00, 0x0403f400, 0x0404000c, 0x0403fff4, 0xfbfc0000, 0xfbfc0400, 0xfbfbfc00, - 0xfbfc0004, 0xfbfbfffc, 0xfbfc0404, 0xfbfbfbfc, 0xfbfc0c0c, 0xfbfbf3f4, 0xfbfc0c00, 0xfbfbf400, - 0xfbfc000c, 0xfbfbfff4, 0x0c0c0000, 0x0c0c0400, 0x0c0bfc00, 0x0c0c0004, 0x0c0bfffc, 0x0c0c0404, - 0x0c0bfbfc, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c0c00, 0x0c0bf400, 0x0c0c000c, 0x0c0bfff4, 0xf3f40000, - 0xf3f40400, 0xf3f3fc00, 0xf3f40004, 0xf3f3fffc, 0xf3f40404, 0xf3f3fbfc, 0xf3f40c0c, 0xf3f3f3f4, - 0xf3f40c00, 0xf3f3f400, 0xf3f4000c, 0xf3f3fff4, 0x0c000000, 0x0c000400, 0x0bfffc00, 0x0c000004, - 0x0bfffffc, 0x0c000404, 0x0bfffbfc, 0x0c000c0c, 0x0bfff3f4, 0x0c000c00, 0x0bfff400, 0x0c00000c, - 0x0bfffff4, 0xf4000000, 0xf4000400, 0xf3fffc00, 0xf4000004, 0xf3fffffc, 0xf4000404, 0xf3fffbfc, - 0xf4000c0c, 0xf3fff3f4, 0xf4000c00, 0xf3fff400, 0xf400000c, 0xf3fffff4, 0x000c0000, 0x000c0400, - 0x000bfc00, 0x000c0004, 0x000bfffc, 0x000c0404, 0x000bfbfc, 0x000c0c0c, 0x000bf3f4, 0x000c0c00, - 0x000bf400, 0x000c000c, 0x000bfff4, 0xfff40000, 0xfff40400, 0xfff3fc00, 0xfff40004, 0xfff3fffc, - 0xfff40404, 0xfff3fbfc, 0xfff40c0c, 0xfff3f3f4, 0xfff40c00, 0xfff3f400, 0xfff4000c, 0xfff3fff4, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, - 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, - 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, - 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, - 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, - 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, - 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, - 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, - 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, - 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, - 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, - 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, - 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, - 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, - 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, - 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, - 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, - 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, - 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, - 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, - 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, - 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, - 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, - 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, - 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, - 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, - 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, - 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, - 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, - 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, - 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, - 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, - 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, - 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, - 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, - 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, - 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, - 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, - 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, - 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, - 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, - 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, - 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, - 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, - 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, - 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, - 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, - 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, - 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, - 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, - 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, - 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, - 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, - 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, - 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, - 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, - 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, - 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, - 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, - 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, - 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, - 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, - 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, - 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, - 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, - 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, - 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, - 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, - 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, - 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, - 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, - 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, - 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, - 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, - 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, - 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, - 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, - 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, - 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, - 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, - 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, - 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, - 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, - 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, - 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 + 0x00000000, 0x00000202, 0xfffffdfe, 0x000002ff, 0xfffffd01, 0xffffff03, 0x000000fd, 0x00000404, + 0xfffffbfc, 0x00000501, 0xfffffaff, 0x00000105, 0xfffffefb, 0x000003fc, 0xfffffc04, 0x000005fe, + 0xfffffa02, 0xfffffe06, 0x000001fa, 0x00000904, 0xfffff6fc, 0x00000409, 0xfffffbf7, 0x00000909, + 0xfffff6f7, 0x00000a01, 0xfffff5ff, 0x0000010a, 0xfffffef6, 0x000007fb, 0xfffff805, 0xfffffb08, + 0x000004f8, 0x00000f09, 0xfffff0f7, 0x0000090f, 0xfffff6f1, 0x00000bfd, 0xfffff403, 0xfffffd0c, + 0x000002f4, 0x00001004, 0xffffeffc, 0x00000410, 0xfffffbf0, 0x00001010, 0xffffeff0, 0x00001200, + 0xffffee00, 0x00000012, 0xffffffee, 0x00000bf4, 0xfffff40c, 0x00000ff7, 0xfffff009, 0xfffff710, + 0x000008f0, 0x00001b0b, 0xffffe4f5, 0x00000b1b, 0xfffff4e5, 0x00001c13, 0xffffe3ed, 0x0000131c, + 0xffffece4, 0x000015fa, 0xffffea06, 0xfffffa16, 0x000005ea, 0x00001d04, 0xffffe2fc, 0x0000041d, + 0xfffffbe3, 0x00001e1e, 0xffffe1e2, 0x000020fe, 0xffffdf02, 0xfffffe21, 0x000001df, 0x000016ee, + 0xffffe912, 0xffffee17, 0x000011e9, 0x00001df1, 0xffffe20f, 0xfffff11e, 0x00000ee2, 0x00002e16, + 0xffffd1ea, 0x0000162e, 0xffffe9d2, 0x00002f0d, 0xffffd0f3, 0x00000d2f, 0xfffff2d1, 0x00003123, + 0xffffcedd, 0x00002331, 0xffffdccf, 0x000028f5, 0xffffd70b, 0xfffff529, 0x00000ad7, 0x00003304, + 0xffffccfc, 0x00000433, 0xfffffbcd, 0x00003636, 0xffffc9ca, 0x000021de, 0xffffde22, 0x000029e3, + 0xffffd61d, 0xffffe32a, 0x00001cd6, 0x00003bfa, 0xffffc406, 0xfffffa3c, 0x000005c4, 0x00004c1b, + 0xffffb3e5, 0x00001b4c, 0xffffe4b4, 0x00004d2b, 0xffffb2d5, 0x00002b4d, 0xffffd4b3, 0x000036e8, + 0xffffc918, 0xffffe837, 0x000017c9, 0x00004f0e, 0xffffb0f2, 0x00000e4f, 0xfffff1b1, 0x0000533f, + 0xffffacc1, 0x00003f53, 0xffffc0ad, 0x000049ec, 0xffffb614, 0xffffec4a, 0x000013b6, 0x00005802, + 0xffffa7fe, 0x00000258, 0xfffffda8, 0x00005d5d, 0xffffa2a3, 0x00003ccc, 0xffffc334, 0xffffcc3d, + 0x000033c3, 0x00007834, 0xffff87cc, 0x00003478, 0xffffcb88, 0x00004ad3, 0xffffb52d, 0xffffd34b, + 0x00002cb5, 0x00007d4b, 0xffff82b5, 0x00004b7d, 0xffffb483, 0x00007a21, 0xffff85df, 0x0000217a, + 0xffffde86, 0x000066f3, 0xffff990d, 0xfffff367, 0x00000c99, 0x00005fd8, 0xffffa028, 0xffffd860, + 0x000027a0, 0x00007ede, 0xffff8122, 0xffffde7f, 0x00002181, 0x000058a7, 0xffffa759, 0x000068b2, + 0xffff974e, 0xffffb269, 0x00004d97, 0x00000c0c, 0xfffff3f4, 0x00001717, 0xffffe8e9, 0x00002a2a, + 0xffffd5d6, 0x00004949, 0xffffb6b7, 0x00000000, 0x02020000, 0xfdfe0000, 0x02ff0000, 0xfd010000, + 0xff030000, 0x00fd0000, 0x00000202, 0x02020202, 0xfdfe0202, 0x02ff0202, 0xfd010202, 0xff030202, + 0x00fd0202, 0xfffffdfe, 0x0201fdfe, 0xfdfdfdfe, 0x02fefdfe, 0xfd00fdfe, 0xff02fdfe, 0x00fcfdfe, + 0x000002ff, 0x020202ff, 0xfdfe02ff, 0x02ff02ff, 0xfd0102ff, 0xff0302ff, 0x00fd02ff, 0xfffffd01, + 0x0201fd01, 0xfdfdfd01, 0x02fefd01, 0xfd00fd01, 0xff02fd01, 0x00fcfd01, 0xffffff03, 0x0201ff03, + 0xfdfdff03, 0x02feff03, 0xfd00ff03, 0xff02ff03, 0x00fcff03, 0x000000fd, 0x020200fd, 0xfdfe00fd, + 0x02ff00fd, 0xfd0100fd, 0xff0300fd, 0x00fd00fd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000303, 0xfffffcfd, 0x000003ff, 0xfffffc01, 0xffffff04, 0x000000fc, 0x00000707, + 0xfffff8f9, 0x00000802, 0xfffff7fe, 0x00000208, 0xfffffdf8, 0x000008fe, 0xfffff702, 0xfffffe09, + 0x000001f7, 0x000005fa, 0xfffffa06, 0x00000d06, 0xfffff2fa, 0x0000060d, 0xfffff9f3, 0x00000d0d, + 0xfffff2f3, 0x00000e01, 0xfffff1ff, 0x0000010e, 0xfffffef2, 0x00000bf8, 0xfffff408, 0xfffff80c, + 0x000007f4, 0x0000170e, 0xffffe8f2, 0x00000e17, 0xfffff1e9, 0x000011fb, 0xffffee05, 0xfffffb12, + 0x000004ee, 0x00001806, 0xffffe7fa, 0x00000618, 0xfffff9e8, 0x00001818, 0xffffe7e8, 0x00001aff, + 0xffffe501, 0xffffff1b, 0x000000e5, 0x000010ef, 0xffffef11, 0x000016f3, 0xffffe90d, 0xfffff317, + 0x00000ce9, 0x00002810, 0xffffd7f0, 0x00001028, 0xffffefd8, 0x0000291c, 0xffffd6e4, 0x00001c29, + 0xffffe3d7, 0x000020f7, 0xffffdf09, 0xfffff721, 0x000008df, 0x00002b06, 0xffffd4fa, 0x0000062b, + 0xfffff9d5, 0x00002e2e, 0xffffd1d2, 0x000031fc, 0xffffce04, 0xfffffc32, 0x000003ce, 0x000021e5, + 0xffffde1b, 0xffffe522, 0x00001ade, 0x00002cea, 0xffffd316, 0xffffea2d, 0x000015d3, 0x00004522, + 0xffffbade, 0x00002245, 0xffffddbb, 0x00004613, 0xffffb9ed, 0x00001346, 0xffffecba, 0x00004935, + 0xffffb6cb, 0x00003549, 0xffffcab7, 0x00003def, 0xffffc211, 0xffffef3e, 0x000010c2, 0x00004d05, + 0xffffb2fb, 0x0000054d, 0xfffffab3, 0x00005252, 0xffffadae, 0x000032cd, 0xffffcd33, 0x00003fd5, + 0xffffc02b, 0xffffd540, 0x00002ac0, 0x000059f6, 0xffffa60a, 0xfffff65a, 0x000009a6, 0x00007229, + 0xffff8dd7, 0x00002972, 0xffffd68e, 0x00007440, 0xffff8bc0, 0x00004074, 0xffffbf8c, 0x000051db, + 0xffffae25, 0xffffdb52, 0x000024ae, 0x00007716, 0xffff88ea, 0x00001677, 0xffffe989, 0x00007c5f, + 0xffff83a1, 0x00005f7c, 0xffffa084, 0x00006ee2, 0xffff911e, 0xffffe26f, 0x00001d91, 0x00005bb2, + 0xffffa44e, 0xffffb25c, 0x00004da4, 0x000070bc, 0xffff8f44, 0xffffbc71, 0x0000438f, 0x00001212, + 0xffffedee, 0x00002222, 0xffffddde, 0x00003f3f, 0xffffc0c1, 0x00006d6d, 0xffff9293, 0x00000000, + 0x03030000, 0xfcfd0000, 0x03ff0000, 0xfc010000, 0xff040000, 0x00fc0000, 0x07070000, 0xf8f90000, + 0x00000303, 0x03030303, 0xfcfd0303, 0x03ff0303, 0xfc010303, 0xff040303, 0x00fc0303, 0x07070303, + 0xf8f90303, 0xfffffcfd, 0x0302fcfd, 0xfcfcfcfd, 0x03fefcfd, 0xfc00fcfd, 0xff03fcfd, 0x00fbfcfd, + 0x0706fcfd, 0xf8f8fcfd, 0x000003ff, 0x030303ff, 0xfcfd03ff, 0x03ff03ff, 0xfc0103ff, 0xff0403ff, + 0x00fc03ff, 0x070703ff, 0xf8f903ff, 0xfffffc01, 0x0302fc01, 0xfcfcfc01, 0x03fefc01, 0xfc00fc01, + 0xff03fc01, 0x00fbfc01, 0x0706fc01, 0xf8f8fc01, 0xffffff04, 0x0302ff04, 0xfcfcff04, 0x03feff04, + 0xfc00ff04, 0xff03ff04, 0x00fbff04, 0x0706ff04, 0xf8f8ff04, 0x000000fc, 0x030300fc, 0xfcfd00fc, + 0x03ff00fc, 0xfc0100fc, 0xff0400fc, 0x00fc00fc, 0x070700fc, 0xf8f900fc, 0x00000707, 0x03030707, + 0xfcfd0707, 0x03ff0707, 0xfc010707, 0xff040707, 0x00fc0707, 0x07070707, 0xf8f90707, 0xfffff8f9, + 0x0302f8f9, 0xfcfcf8f9, 0x03fef8f9, 0xfc00f8f9, 0xff03f8f9, 0x00fbf8f9, 0x0706f8f9, 0xf8f8f8f9, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000404, 0xfffffbfc, 0x000004ff, 0xfffffb01, 0xffffff05, 0x000000fb, 0x00000a03, + 0xfffff5fd, 0x0000030a, 0xfffffcf6, 0x00000909, 0xfffff6f7, 0x000006f9, 0xfffff907, 0x00000bfd, + 0xfffff403, 0xfffffd0c, 0x000002f4, 0x00001108, 0xffffeef8, 0x00000811, 0xfffff7ef, 0x00001111, + 0xffffeeef, 0x00001301, 0xffffecff, 0x00000113, 0xfffffeed, 0x00000ff5, 0xfffff00b, 0xfffff510, + 0x00000af0, 0x000016fa, 0xffffe906, 0xfffffa17, 0x000005e9, 0x00001f12, 0xffffe0ee, 0x0000121f, + 0xffffede1, 0x00002008, 0xffffdff8, 0x00000820, 0xfffff7e0, 0x00002121, 0xffffdedf, 0x000023ff, + 0xffffdc01, 0xffffff24, 0x000000dc, 0x000016e9, 0xffffe917, 0x00001eef, 0xffffe111, 0xffffef1f, + 0x000010e1, 0x00003615, 0xffffc9eb, 0x00001536, 0xffffeaca, 0x00003725, 0xffffc8db, 0x00002537, + 0xffffdac9, 0x00002bf4, 0xffffd40c, 0xfffff42c, 0x00000bd4, 0x00003908, 0xffffc6f8, 0x00000839, + 0xfffff7c7, 0x00003d3d, 0xffffc2c3, 0x000041fb, 0xffffbe05, 0xfffffb42, 0x000004be, 0x00002cdc, + 0xffffd324, 0xffffdc2d, 0x000023d3, 0x00003be3, 0xffffc41d, 0xffffe33c, 0x00001cc4, 0x00005c2d, + 0xffffa3d3, 0x00002d5c, 0xffffd2a4, 0x00005d19, 0xffffa2e7, 0x0000195d, 0xffffe6a3, 0x00006147, + 0xffff9eb9, 0x00004761, 0xffffb89f, 0x000052ea, 0xffffad16, 0xffffea53, 0x000015ad, 0x00006607, + 0xffff99f9, 0x00000766, 0xfffff89a, 0x00006d6d, 0xffff9293, 0x000043bc, 0xffffbc44, 0x000054c7, + 0xffffab39, 0xffffc755, 0x000038ab, 0x000077f3, 0xffff880d, 0xfffff378, 0x00000c88, 0x00006dcf, + 0xffff9231, 0xffffcf6e, 0x00003092, 0x00007a98, 0xffff8568, 0xffff987b, 0x00006785, 0x00001818, + 0xffffe7e8, 0x00002e2e, 0xffffd1d2, 0x00005454, 0xffffabac, 0x00000000, 0x04040000, 0xfbfc0000, + 0x04ff0000, 0xfb010000, 0xff050000, 0x00fb0000, 0x0a030000, 0xf5fd0000, 0x030a0000, 0x00000404, + 0x04040404, 0xfbfc0404, 0x04ff0404, 0xfb010404, 0xff050404, 0x00fb0404, 0x0a030404, 0xf5fd0404, + 0x030a0404, 0xfffffbfc, 0x0403fbfc, 0xfbfbfbfc, 0x04fefbfc, 0xfb00fbfc, 0xff04fbfc, 0x00fafbfc, + 0x0a02fbfc, 0xf5fcfbfc, 0x0309fbfc, 0x000004ff, 0x040404ff, 0xfbfc04ff, 0x04ff04ff, 0xfb0104ff, + 0xff0504ff, 0x00fb04ff, 0x0a0304ff, 0xf5fd04ff, 0x030a04ff, 0xfffffb01, 0x0403fb01, 0xfbfbfb01, + 0x04fefb01, 0xfb00fb01, 0xff04fb01, 0x00fafb01, 0x0a02fb01, 0xf5fcfb01, 0x0309fb01, 0xffffff05, + 0x0403ff05, 0xfbfbff05, 0x04feff05, 0xfb00ff05, 0xff04ff05, 0x00faff05, 0x0a02ff05, 0xf5fcff05, + 0x0309ff05, 0x000000fb, 0x040400fb, 0xfbfc00fb, 0x04ff00fb, 0xfb0100fb, 0xff0500fb, 0x00fb00fb, + 0x0a0300fb, 0xf5fd00fb, 0x030a00fb, 0x00000a03, 0x04040a03, 0xfbfc0a03, 0x04ff0a03, 0xfb010a03, + 0xff050a03, 0x00fb0a03, 0x0a030a03, 0xf5fd0a03, 0x030a0a03, 0xfffff5fd, 0x0403f5fd, 0xfbfbf5fd, + 0x04fef5fd, 0xfb00f5fd, 0xff04f5fd, 0x00faf5fd, 0x0a02f5fd, 0xf5fcf5fd, 0x0309f5fd, 0x0000030a, + 0x0404030a, 0xfbfc030a, 0x04ff030a, 0xfb01030a, 0xff05030a, 0x00fb030a, 0x0a03030a, 0xf5fd030a, + 0x030a030a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000505, 0xfffffafb, 0x000006fe, 0xfffff902, 0xfffffe07, 0x000001f9, 0x00000b0b, + 0xfffff4f5, 0x00000d03, 0xfffff2fd, 0x0000030d, 0xfffffcf3, 0x000008f7, 0xfffff709, 0x00000efc, + 0xfffff104, 0xfffffc0f, 0x000003f1, 0x0000160b, 0xffffe9f5, 0x00000b16, 0xfffff4ea, 0x00001515, + 0xffffeaeb, 0x00001802, 0xffffe7fe, 0x00000218, 0xfffffde8, 0x000013f2, 0xffffec0e, 0xfffff214, + 0x00000dec, 0x00002617, 0xffffd9e9, 0x00001726, 0xffffe8da, 0x00001cf8, 0xffffe308, 0xfffff81d, + 0x000007e3, 0x0000270b, 0xffffd8f5, 0x00000b27, 0xfffff4d9, 0x00002929, 0xffffd6d7, 0x00002cff, + 0xffffd301, 0xffffff2d, 0x000000d3, 0x00001ce3, 0xffffe31d, 0x000026ea, 0xffffd916, 0xffffea27, + 0x000015d9, 0x0000431b, 0xffffbce5, 0x00001b43, 0xffffe4bd, 0x0000452f, 0xffffbad1, 0x00002f45, + 0xffffd0bb, 0x000037f1, 0xffffc80f, 0xfffff138, 0x00000ec8, 0x0000470b, 0xffffb8f5, 0x00000b47, + 0xfffff4b9, 0x00004c4c, 0xffffb3b4, 0x000052fa, 0xffffad06, 0xfffffa53, 0x000005ad, 0x000038d3, + 0xffffc72d, 0xffffd339, 0x00002cc7, 0x00004adc, 0xffffb524, 0xffffdc4b, 0x000023b5, 0x00007338, + 0xffff8cc8, 0x00003873, 0xffffc78d, 0x0000751f, 0xffff8ae1, 0x00001f75, 0xffffe08b, 0x00007a58, + 0xffff85a8, 0x0000587a, 0xffffa786, 0x000067e4, 0xffff981c, 0xffffe468, 0x00001b98, 0x000054ab, + 0xffffab55, 0x000069b8, 0xffff9648, 0xffffb86a, 0x00004796, 0x00001e1e, 0xffffe1e2, 0x00003a3a, + 0xffffc5c6, 0x00006969, 0xffff9697, 0x00000000, 0x05050000, 0xfafb0000, 0x06fe0000, 0xf9020000, + 0xfe070000, 0x01f90000, 0x0b0b0000, 0xf4f50000, 0x0d030000, 0xf2fd0000, 0x00000505, 0x05050505, + 0xfafb0505, 0x06fe0505, 0xf9020505, 0xfe070505, 0x01f90505, 0x0b0b0505, 0xf4f50505, 0x0d030505, + 0xf2fd0505, 0xfffffafb, 0x0504fafb, 0xfafafafb, 0x06fdfafb, 0xf901fafb, 0xfe06fafb, 0x01f8fafb, + 0x0b0afafb, 0xf4f4fafb, 0x0d02fafb, 0xf2fcfafb, 0x000006fe, 0x050506fe, 0xfafb06fe, 0x06fe06fe, + 0xf90206fe, 0xfe0706fe, 0x01f906fe, 0x0b0b06fe, 0xf4f506fe, 0x0d0306fe, 0xf2fd06fe, 0xfffff902, + 0x0504f902, 0xfafaf902, 0x06fdf902, 0xf901f902, 0xfe06f902, 0x01f8f902, 0x0b0af902, 0xf4f4f902, + 0x0d02f902, 0xf2fcf902, 0xfffffe07, 0x0504fe07, 0xfafafe07, 0x06fdfe07, 0xf901fe07, 0xfe06fe07, + 0x01f8fe07, 0x0b0afe07, 0xf4f4fe07, 0x0d02fe07, 0xf2fcfe07, 0x000001f9, 0x050501f9, 0xfafb01f9, + 0x06fe01f9, 0xf90201f9, 0xfe0701f9, 0x01f901f9, 0x0b0b01f9, 0xf4f501f9, 0x0d0301f9, 0xf2fd01f9, + 0x00000b0b, 0x05050b0b, 0xfafb0b0b, 0x06fe0b0b, 0xf9020b0b, 0xfe070b0b, 0x01f90b0b, 0x0b0b0b0b, + 0xf4f50b0b, 0x0d030b0b, 0xf2fd0b0b, 0xfffff4f5, 0x0504f4f5, 0xfafaf4f5, 0x06fdf4f5, 0xf901f4f5, + 0xfe06f4f5, 0x01f8f4f5, 0x0b0af4f5, 0xf4f4f4f5, 0x0d02f4f5, 0xf2fcf4f5, 0x00000d03, 0x05050d03, + 0xfafb0d03, 0x06fe0d03, 0xf9020d03, 0xfe070d03, 0x01f90d03, 0x0b0b0d03, 0xf4f50d03, 0x0d030d03, + 0xf2fd0d03, 0xfffff2fd, 0x0504f2fd, 0xfafaf2fd, 0x06fdf2fd, 0xf901f2fd, 0xfe06f2fd, 0x01f8f2fd, + 0x0b0af2fd, 0xf4f4f2fd, 0x0d02f2fd, 0xf2fcf2fd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000606, 0xfffff9fa, 0x000007fe, 0xfffff802, 0xfffffe08, 0x000001f8, 0x00000d0d, + 0xfffff2f3, 0x00000f04, 0xfffff0fc, 0x0000040f, 0xfffffbf1, 0x00000af5, 0xfffff50b, 0x000011fb, + 0xffffee05, 0xfffffb12, 0x000004ee, 0x00001a0d, 0xffffe5f3, 0x00000d1a, 0xfffff2e6, 0x00001a1a, + 0xffffe5e6, 0x00001d02, 0xffffe2fe, 0x0000021d, 0xfffffde3, 0x000017f0, 0xffffe810, 0xfffff018, + 0x00000fe8, 0x00002e1c, 0xffffd1e4, 0x00001c2e, 0xffffe3d2, 0x000022f7, 0xffffdd09, 0xfffff723, + 0x000008dd, 0x00002f0d, 0xffffd0f3, 0x00000d2f, 0xfffff2d1, 0x00003131, 0xffffcecf, 0x000035ff, + 0xffffca01, 0xffffff36, 0x000000ca, 0x000022dd, 0xffffdd23, 0x00002ee6, 0xffffd11a, 0xffffe62f, + 0x000019d1, 0x00005120, 0xffffaee0, 0x00002051, 0xffffdfaf, 0x00005338, 0xffffacc8, 0x00003853, + 0xffffc7ad, 0x000042ee, 0xffffbd12, 0xffffee43, 0x000011bd, 0x0000560d, 0xffffa9f3, 0x00000d56, + 0xfffff2aa, 0x00005b5b, 0xffffa4a5, 0x000062f9, 0xffff9d07, 0xfffff963, 0x0000069d, 0x000043ca, + 0xffffbc36, 0xffffca44, 0x000035bc, 0x000059d4, 0xffffa62c, 0xffffd45a, 0x00002ba6, 0x00007bdf, + 0xffff8421, 0xffffdf7c, 0x00002084, 0x00006699, 0xffff9967, 0x00007eaa, 0xffff8156, 0xffffaa7f, + 0x00005581, 0x00002525, 0xffffdadb, 0x00004545, 0xffffbabb, 0x00000000, 0x06060000, 0xf9fa0000, + 0x07fe0000, 0xf8020000, 0xfe080000, 0x01f80000, 0x0d0d0000, 0xf2f30000, 0x0f040000, 0xf0fc0000, + 0x040f0000, 0x00000606, 0x06060606, 0xf9fa0606, 0x07fe0606, 0xf8020606, 0xfe080606, 0x01f80606, + 0x0d0d0606, 0xf2f30606, 0x0f040606, 0xf0fc0606, 0x040f0606, 0xfffff9fa, 0x0605f9fa, 0xf9f9f9fa, + 0x07fdf9fa, 0xf801f9fa, 0xfe07f9fa, 0x01f7f9fa, 0x0d0cf9fa, 0xf2f2f9fa, 0x0f03f9fa, 0xf0fbf9fa, + 0x040ef9fa, 0x000007fe, 0x060607fe, 0xf9fa07fe, 0x07fe07fe, 0xf80207fe, 0xfe0807fe, 0x01f807fe, + 0x0d0d07fe, 0xf2f307fe, 0x0f0407fe, 0xf0fc07fe, 0x040f07fe, 0xfffff802, 0x0605f802, 0xf9f9f802, + 0x07fdf802, 0xf801f802, 0xfe07f802, 0x01f7f802, 0x0d0cf802, 0xf2f2f802, 0x0f03f802, 0xf0fbf802, + 0x040ef802, 0xfffffe08, 0x0605fe08, 0xf9f9fe08, 0x07fdfe08, 0xf801fe08, 0xfe07fe08, 0x01f7fe08, + 0x0d0cfe08, 0xf2f2fe08, 0x0f03fe08, 0xf0fbfe08, 0x040efe08, 0x000001f8, 0x060601f8, 0xf9fa01f8, + 0x07fe01f8, 0xf80201f8, 0xfe0801f8, 0x01f801f8, 0x0d0d01f8, 0xf2f301f8, 0x0f0401f8, 0xf0fc01f8, + 0x040f01f8, 0x00000d0d, 0x06060d0d, 0xf9fa0d0d, 0x07fe0d0d, 0xf8020d0d, 0xfe080d0d, 0x01f80d0d, + 0x0d0d0d0d, 0xf2f30d0d, 0x0f040d0d, 0xf0fc0d0d, 0x040f0d0d, 0xfffff2f3, 0x0605f2f3, 0xf9f9f2f3, + 0x07fdf2f3, 0xf801f2f3, 0xfe07f2f3, 0x01f7f2f3, 0x0d0cf2f3, 0xf2f2f2f3, 0x0f03f2f3, 0xf0fbf2f3, + 0x040ef2f3, 0x00000f04, 0x06060f04, 0xf9fa0f04, 0x07fe0f04, 0xf8020f04, 0xfe080f04, 0x01f80f04, + 0x0d0d0f04, 0xf2f30f04, 0x0f040f04, 0xf0fc0f04, 0x040f0f04, 0xfffff0fc, 0x0605f0fc, 0xf9f9f0fc, + 0x07fdf0fc, 0xf801f0fc, 0xfe07f0fc, 0x01f7f0fc, 0x0d0cf0fc, 0xf2f2f0fc, 0x0f03f0fc, 0xf0fbf0fc, + 0x040ef0fc, 0x0000040f, 0x0606040f, 0xf9fa040f, 0x07fe040f, 0xf802040f, 0xfe08040f, 0x01f8040f, + 0x0d0d040f, 0xf2f3040f, 0x0f04040f, 0xf0fc040f, 0x040f040f, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000707, 0xfffff8f9, 0x000009fd, 0xfffff603, 0xfffffd0a, 0x000002f6, 0x00001010, + 0xffffeff0, 0x00001205, 0xffffedfb, 0x00000512, 0xfffffaee, 0x00000cf3, 0xfffff30d, 0x000014fa, + 0xffffeb06, 0xfffffa15, 0x000005eb, 0x00001e0f, 0xffffe1f1, 0x00000f1e, 0xfffff0e2, 0x00001e1e, + 0xffffe1e2, 0x00002202, 0xffffddfe, 0x00000222, 0xfffffdde, 0x00001bed, 0xffffe413, 0xffffed1c, + 0x000012e4, 0x00003620, 0xffffc9e0, 0x00002036, 0xffffdfca, 0x000028f5, 0xffffd70b, 0xfffff529, + 0x00000ad7, 0x0000370f, 0xffffc8f1, 0x00000f37, 0xfffff0c9, 0x00003939, 0xffffc6c7, 0x00003eff, + 0xffffc101, 0xffffff3f, 0x000000c1, 0x000027d8, 0xffffd828, 0x000036e2, 0xffffc91e, 0xffffe237, + 0x00001dc9, 0x00005e25, 0xffffa1db, 0x0000255e, 0xffffdaa2, 0x00006041, 0xffff9fbf, 0x00004160, + 0xffffbea0, 0x00004deb, 0xffffb215, 0xffffeb4e, 0x000014b2, 0x0000640f, 0xffff9bf1, 0x00000f64, + 0xfffff09c, 0x00006a6a, 0xffff9596, 0x000073f8, 0xffff8c08, 0xfffff874, 0x0000078c, 0x00004ec1, + 0xffffb13f, 0xffffc14f, 0x00003eb1, 0x000068cd, 0xffff9733, 0xffffcd69, 0x00003297, 0x00007788, + 0xffff8878, 0x00002b2b, 0xffffd4d5, 0x00005050, 0xffffafb0, 0x00000000, 0x07070000, 0xf8f90000, + 0x09fd0000, 0xf6030000, 0xfd0a0000, 0x02f60000, 0x10100000, 0xeff00000, 0x12050000, 0xedfb0000, + 0x05120000, 0x00000707, 0x07070707, 0xf8f90707, 0x09fd0707, 0xf6030707, 0xfd0a0707, 0x02f60707, + 0x10100707, 0xeff00707, 0x12050707, 0xedfb0707, 0x05120707, 0xfffff8f9, 0x0706f8f9, 0xf8f8f8f9, + 0x09fcf8f9, 0xf602f8f9, 0xfd09f8f9, 0x02f5f8f9, 0x100ff8f9, 0xefeff8f9, 0x1204f8f9, 0xedfaf8f9, + 0x0511f8f9, 0x000009fd, 0x070709fd, 0xf8f909fd, 0x09fd09fd, 0xf60309fd, 0xfd0a09fd, 0x02f609fd, + 0x101009fd, 0xeff009fd, 0x120509fd, 0xedfb09fd, 0x051209fd, 0xfffff603, 0x0706f603, 0xf8f8f603, + 0x09fcf603, 0xf602f603, 0xfd09f603, 0x02f5f603, 0x100ff603, 0xefeff603, 0x1204f603, 0xedfaf603, + 0x0511f603, 0xfffffd0a, 0x0706fd0a, 0xf8f8fd0a, 0x09fcfd0a, 0xf602fd0a, 0xfd09fd0a, 0x02f5fd0a, + 0x100ffd0a, 0xefeffd0a, 0x1204fd0a, 0xedfafd0a, 0x0511fd0a, 0x000002f6, 0x070702f6, 0xf8f902f6, + 0x09fd02f6, 0xf60302f6, 0xfd0a02f6, 0x02f602f6, 0x101002f6, 0xeff002f6, 0x120502f6, 0xedfb02f6, + 0x051202f6, 0x00001010, 0x07071010, 0xf8f91010, 0x09fd1010, 0xf6031010, 0xfd0a1010, 0x02f61010, + 0x10101010, 0xeff01010, 0x12051010, 0xedfb1010, 0x05121010, 0xffffeff0, 0x0706eff0, 0xf8f8eff0, + 0x09fceff0, 0xf602eff0, 0xfd09eff0, 0x02f5eff0, 0x100feff0, 0xefefeff0, 0x1204eff0, 0xedfaeff0, + 0x0511eff0, 0x00001205, 0x07071205, 0xf8f91205, 0x09fd1205, 0xf6031205, 0xfd0a1205, 0x02f61205, + 0x10101205, 0xeff01205, 0x12051205, 0xedfb1205, 0x05121205, 0xffffedfb, 0x0706edfb, 0xf8f8edfb, + 0x09fcedfb, 0xf602edfb, 0xfd09edfb, 0x02f5edfb, 0x100fedfb, 0xefefedfb, 0x1204edfb, 0xedfaedfb, + 0x0511edfb, 0x00000512, 0x07070512, 0xf8f90512, 0x09fd0512, 0xf6030512, 0xfd0a0512, 0x02f60512, + 0x10100512, 0xeff00512, 0x12050512, 0xedfb0512, 0x05120512, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000808, 0xfffff7f8, 0x00000afd, 0xfffff503, 0xfffffd0b, 0x000002f5, 0x00001212, + 0xffffedee, 0x00001405, 0xffffebfb, 0x00000514, 0xfffffaec, 0x00000ef1, 0xfffff10f, 0x000017f9, + 0xffffe807, 0xfffff918, 0x000006e8, 0x00002311, 0xffffdcef, 0x00001123, 0xffffeedd, 0x00002222, + 0xffffddde, 0x00002603, 0xffffd9fd, 0x00000326, 0xfffffcda, 0x00001fea, 0xffffe016, 0xffffea20, + 0x000015e0, 0x00003d25, 0xffffc2db, 0x0000253d, 0xffffdac3, 0x00002ef3, 0xffffd10d, 0xfffff32f, + 0x00000cd1, 0x00003f11, 0xffffc0ef, 0x0000113f, 0xffffeec1, 0x00004141, 0xffffbebf, 0x000047ff, + 0xffffb801, 0xffffff48, 0x000000b8, 0x00002dd2, 0xffffd22e, 0x00003edd, 0xffffc123, 0xffffdd3f, + 0x000022c1, 0x00006b2b, 0xffff94d5, 0x00002b6b, 0xffffd495, 0x00006e4b, 0xffff91b5, 0x00004b6e, + 0xffffb492, 0x000058e8, 0xffffa718, 0xffffe859, 0x000017a7, 0x00007211, 0xffff8def, 0x00001172, + 0xffffee8e, 0x00007979, 0xffff8687, 0x00005ab8, 0xffffa548, 0xffffb85b, 0x000047a5, 0x000077c6, + 0xffff883a, 0xffffc678, 0x00003988, 0x00003131, 0xffffcecf, 0x00005c5c, 0xffffa3a4, 0x00000000, + 0x08080000, 0xf7f80000, 0x0afd0000, 0xf5030000, 0xfd0b0000, 0x02f50000, 0x12120000, 0xedee0000, + 0x14050000, 0xebfb0000, 0x05140000, 0x00000808, 0x08080808, 0xf7f80808, 0x0afd0808, 0xf5030808, + 0xfd0b0808, 0x02f50808, 0x12120808, 0xedee0808, 0x14050808, 0xebfb0808, 0x05140808, 0xfffff7f8, + 0x0807f7f8, 0xf7f7f7f8, 0x0afcf7f8, 0xf502f7f8, 0xfd0af7f8, 0x02f4f7f8, 0x1211f7f8, 0xededf7f8, + 0x1404f7f8, 0xebfaf7f8, 0x0513f7f8, 0x00000afd, 0x08080afd, 0xf7f80afd, 0x0afd0afd, 0xf5030afd, + 0xfd0b0afd, 0x02f50afd, 0x12120afd, 0xedee0afd, 0x14050afd, 0xebfb0afd, 0x05140afd, 0xfffff503, + 0x0807f503, 0xf7f7f503, 0x0afcf503, 0xf502f503, 0xfd0af503, 0x02f4f503, 0x1211f503, 0xededf503, + 0x1404f503, 0xebfaf503, 0x0513f503, 0xfffffd0b, 0x0807fd0b, 0xf7f7fd0b, 0x0afcfd0b, 0xf502fd0b, + 0xfd0afd0b, 0x02f4fd0b, 0x1211fd0b, 0xededfd0b, 0x1404fd0b, 0xebfafd0b, 0x0513fd0b, 0x000002f5, + 0x080802f5, 0xf7f802f5, 0x0afd02f5, 0xf50302f5, 0xfd0b02f5, 0x02f502f5, 0x121202f5, 0xedee02f5, + 0x140502f5, 0xebfb02f5, 0x051402f5, 0x00001212, 0x08081212, 0xf7f81212, 0x0afd1212, 0xf5031212, + 0xfd0b1212, 0x02f51212, 0x12121212, 0xedee1212, 0x14051212, 0xebfb1212, 0x05141212, 0xffffedee, + 0x0807edee, 0xf7f7edee, 0x0afcedee, 0xf502edee, 0xfd0aedee, 0x02f4edee, 0x1211edee, 0xedededee, + 0x1404edee, 0xebfaedee, 0x0513edee, 0x00001405, 0x08081405, 0xf7f81405, 0x0afd1405, 0xf5031405, + 0xfd0b1405, 0x02f51405, 0x12121405, 0xedee1405, 0x14051405, 0xebfb1405, 0x05141405, 0xffffebfb, + 0x0807ebfb, 0xf7f7ebfb, 0x0afcebfb, 0xf502ebfb, 0xfd0aebfb, 0x02f4ebfb, 0x1211ebfb, 0xededebfb, + 0x1404ebfb, 0xebfaebfb, 0x0513ebfb, 0x00000514, 0x08080514, 0xf7f80514, 0x0afd0514, 0xf5030514, + 0xfd0b0514, 0x02f50514, 0x12120514, 0xedee0514, 0x14050514, 0xebfb0514, 0x05140514, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000909, 0xfffff6f7, 0x00000bfd, 0xfffff403, 0xfffffd0c, 0x000002f4, 0x00001414, + 0xffffebec, 0x00001706, 0xffffe8fa, 0x00000617, 0xfffff9e9, 0x000010ef, 0xffffef11, 0x00001af9, + 0xffffe507, 0xfffff91b, 0x000006e5, 0x00002713, 0xffffd8ed, 0x00001327, 0xffffecd9, 0x00002727, + 0xffffd8d9, 0x00002b03, 0xffffd4fd, 0x0000032b, 0xfffffcd5, 0x000023e8, 0xffffdc18, 0xffffe824, + 0x000017dc, 0x0000452a, 0xffffbad6, 0x00002a45, 0xffffd5bb, 0x000034f2, 0xffffcb0e, 0xfffff235, + 0x00000dcb, 0x00004713, 0xffffb8ed, 0x00001347, 0xffffecb9, 0x00004949, 0xffffb6b7, 0x00004ffe, + 0xffffb002, 0xfffffe50, 0x000001b0, 0x000033cc, 0xffffcc34, 0x000045d9, 0xffffba27, 0xffffd946, + 0x000026ba, 0x00007930, 0xffff86d0, 0x00003079, 0xffffcf87, 0x00007c54, 0xffff83ac, 0x0000547c, + 0xffffab84, 0x000063e5, 0xffff9c1b, 0xffffe564, 0x00001a9c, 0x000065af, 0xffff9a51, 0xffffaf66, + 0x0000509a, 0x00003737, 0xffffc8c9, 0x00006868, 0xffff9798, 0x00000000, 0x09090000, 0xf6f70000, + 0x0bfd0000, 0xf4030000, 0xfd0c0000, 0x02f40000, 0x14140000, 0xebec0000, 0x17060000, 0xe8fa0000, + 0x06170000, 0xf9e90000, 0x00000909, 0x09090909, 0xf6f70909, 0x0bfd0909, 0xf4030909, 0xfd0c0909, + 0x02f40909, 0x14140909, 0xebec0909, 0x17060909, 0xe8fa0909, 0x06170909, 0xf9e90909, 0xfffff6f7, + 0x0908f6f7, 0xf6f6f6f7, 0x0bfcf6f7, 0xf402f6f7, 0xfd0bf6f7, 0x02f3f6f7, 0x1413f6f7, 0xebebf6f7, + 0x1705f6f7, 0xe8f9f6f7, 0x0616f6f7, 0xf9e8f6f7, 0x00000bfd, 0x09090bfd, 0xf6f70bfd, 0x0bfd0bfd, + 0xf4030bfd, 0xfd0c0bfd, 0x02f40bfd, 0x14140bfd, 0xebec0bfd, 0x17060bfd, 0xe8fa0bfd, 0x06170bfd, + 0xf9e90bfd, 0xfffff403, 0x0908f403, 0xf6f6f403, 0x0bfcf403, 0xf402f403, 0xfd0bf403, 0x02f3f403, + 0x1413f403, 0xebebf403, 0x1705f403, 0xe8f9f403, 0x0616f403, 0xf9e8f403, 0xfffffd0c, 0x0908fd0c, + 0xf6f6fd0c, 0x0bfcfd0c, 0xf402fd0c, 0xfd0bfd0c, 0x02f3fd0c, 0x1413fd0c, 0xebebfd0c, 0x1705fd0c, + 0xe8f9fd0c, 0x0616fd0c, 0xf9e8fd0c, 0x000002f4, 0x090902f4, 0xf6f702f4, 0x0bfd02f4, 0xf40302f4, + 0xfd0c02f4, 0x02f402f4, 0x141402f4, 0xebec02f4, 0x170602f4, 0xe8fa02f4, 0x061702f4, 0xf9e902f4, + 0x00001414, 0x09091414, 0xf6f71414, 0x0bfd1414, 0xf4031414, 0xfd0c1414, 0x02f41414, 0x14141414, + 0xebec1414, 0x17061414, 0xe8fa1414, 0x06171414, 0xf9e91414, 0xffffebec, 0x0908ebec, 0xf6f6ebec, + 0x0bfcebec, 0xf402ebec, 0xfd0bebec, 0x02f3ebec, 0x1413ebec, 0xebebebec, 0x1705ebec, 0xe8f9ebec, + 0x0616ebec, 0xf9e8ebec, 0x00001706, 0x09091706, 0xf6f71706, 0x0bfd1706, 0xf4031706, 0xfd0c1706, + 0x02f41706, 0x14141706, 0xebec1706, 0x17061706, 0xe8fa1706, 0x06171706, 0xf9e91706, 0xffffe8fa, + 0x0908e8fa, 0xf6f6e8fa, 0x0bfce8fa, 0xf402e8fa, 0xfd0be8fa, 0x02f3e8fa, 0x1413e8fa, 0xebebe8fa, + 0x1705e8fa, 0xe8f9e8fa, 0x0616e8fa, 0xf9e8e8fa, 0x00000617, 0x09090617, 0xf6f70617, 0x0bfd0617, + 0xf4030617, 0xfd0c0617, 0x02f40617, 0x14140617, 0xebec0617, 0x17060617, 0xe8fa0617, 0x06170617, + 0xf9e90617, 0xfffff9e9, 0x0908f9e9, 0xf6f6f9e9, 0x0bfcf9e9, 0xf402f9e9, 0xfd0bf9e9, 0x02f3f9e9, + 0x1413f9e9, 0xebebf9e9, 0x1705f9e9, 0xe8f9f9e9, 0x0616f9e9, 0xf9e8f9e9, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, + 0xfffffbfc, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x000003fc, 0xfffffc04, 0x000005fe, + 0xfffffa02, 0xfffffe06, 0x000001fa, 0x00000804, 0xfffff7fc, 0x00000408, 0xfffffbf8, 0x00000808, + 0xfffff7f8, 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x000007fc, 0xfffff804, 0xfffffc08, + 0x000003f8, 0x00000e08, 0xfffff1f8, 0x0000080e, 0xfffff7f2, 0x00000bfe, 0xfffff402, 0xfffffe0c, + 0x000001f4, 0x00001004, 0xffffeffc, 0x00000410, 0xfffffbf0, 0x00001010, 0xffffeff0, 0x00001200, + 0xffffee00, 0x00000012, 0xffffffee, 0x00000bf4, 0xfffff40c, 0x00000ff8, 0xfffff008, 0xfffff810, + 0x000007f0, 0x00001a0a, 0xffffe5f6, 0x00000a1a, 0xfffff5e6, 0x00001c12, 0xffffe3ee, 0x0000121c, + 0xffffede4, 0x000015fa, 0xffffea06, 0xfffffa16, 0x000005ea, 0x00001c04, 0xffffe3fc, 0x0000041c, + 0xfffffbe4, 0x00001e1e, 0xffffe1e2, 0x00001ffe, 0xffffe002, 0xfffffe20, 0x000001e0, 0x000015ee, + 0xffffea12, 0xffffee16, 0x000011ea, 0x00001df2, 0xffffe20e, 0xfffff21e, 0x00000de2, 0x00002e16, + 0xffffd1ea, 0x0000162e, 0xffffe9d2, 0x00002e0c, 0xffffd1f4, 0x00000c2e, 0xfffff3d2, 0x00003022, + 0xffffcfde, 0x00002230, 0xffffddd0, 0x000027f6, 0xffffd80a, 0xfffff628, 0x000009d8, 0x00003204, + 0xffffcdfc, 0x00000432, 0xfffffbce, 0x00003636, 0xffffc9ca, 0x000021de, 0xffffde22, 0x000029e4, + 0xffffd61c, 0xffffe42a, 0x00001bd6, 0x00003bfa, 0xffffc406, 0xfffffa3c, 0x000005c4, 0x00004c1a, + 0xffffb3e6, 0x00001a4c, 0xffffe5b4, 0x00004c2a, 0xffffb3d6, 0x00002a4c, 0xffffd5b4, 0x000035e8, + 0xffffca18, 0xffffe836, 0x000017ca, 0x00004e0e, 0xffffb1f2, 0x00000e4e, 0xfffff1b2, 0x0000523e, + 0xffffadc2, 0x00003e52, 0xffffc1ae, 0x000049ec, 0xffffb614, 0xffffec4a, 0x000013b6, 0x00005802, + 0xffffa7fe, 0x00000258, 0xfffffda8, 0x00005c5c, 0xffffa3a4, 0x00003bcc, 0xffffc434, 0xffffcc3c, + 0x000033c4, 0x00007634, 0xffff89cc, 0x00003476, 0xffffcb8a, 0x000049d4, 0xffffb62c, 0xffffd44a, + 0x00002bb6, 0x0000764a, 0xffff89b6, 0x00004a76, 0xffffb58a, 0x00007620, 0xffff89e0, 0x00002076, + 0xffffdf8a, 0x000065f4, 0xffff9a0c, 0xfffff466, 0x00000b9a, 0x00005fd8, 0xffffa028, 0xffffd860, + 0x000027a0, 0x000075de, 0xffff8a22, 0xffffde76, 0x0000218a, 0x000057a8, 0xffffa858, 0x000067b2, + 0xffff984e, 0xffffb268, 0x00004d98, 0x00000c0c, 0xfffff3f4, 0x00001616, 0xffffe9ea, 0x00002a2a, + 0xffffd5d6, 0x00004848, 0xffffb7b8, 0x00000000, 0x02020000, 0xfdfe0000, 0x02000000, 0xfe000000, + 0x00020000, 0xfffe0000, 0x00000202, 0x02020202, 0xfdfe0202, 0x02000202, 0xfe000202, 0x00020202, + 0xfffe0202, 0xfffffdfe, 0x0201fdfe, 0xfdfdfdfe, 0x01fffdfe, 0xfdfffdfe, 0x0001fdfe, 0xfffdfdfe, + 0x00000200, 0x02020200, 0xfdfe0200, 0x02000200, 0xfe000200, 0x00020200, 0xfffe0200, 0xfffffe00, + 0x0201fe00, 0xfdfdfe00, 0x01fffe00, 0xfdfffe00, 0x0001fe00, 0xfffdfe00, 0x00000002, 0x02020002, + 0xfdfe0002, 0x02000002, 0xfe000002, 0x00020002, 0xfffe0002, 0xfffffffe, 0x0201fffe, 0xfdfdfffe, + 0x01fffffe, 0xfdfffffe, 0x0001fffe, 0xfffdfffe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000303, 0xfffffcfd, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, + 0xfffff9fa, 0x00000903, 0xfffff6fd, 0x00000309, 0xfffffcf7, 0x000008fd, 0xfffff703, 0xfffffd09, + 0x000002f7, 0x000005fa, 0xfffffa06, 0x00000c06, 0xfffff3fa, 0x0000060c, 0xfffff9f4, 0x00000c0c, + 0xfffff3f4, 0x00000f00, 0xfffff100, 0x0000000f, 0xfffffff1, 0x00000bf7, 0xfffff409, 0xfffff70c, + 0x000008f4, 0x0000180f, 0xffffe7f1, 0x00000f18, 0xfffff0e8, 0x000011fa, 0xffffee06, 0xfffffa12, + 0x000005ee, 0x00001806, 0xffffe7fa, 0x00000618, 0xfffff9e8, 0x00001818, 0xffffe7e8, 0x00001b00, + 0xffffe500, 0x0000001b, 0xffffffe5, 0x000011ee, 0xffffee12, 0x000017f4, 0xffffe80c, 0xfffff418, + 0x00000be8, 0x0000270f, 0xffffd8f1, 0x00000f27, 0xfffff0d9, 0x00002a1b, 0xffffd5e5, 0x00001b2a, + 0xffffe4d6, 0x000020f7, 0xffffdf09, 0xfffff721, 0x000008df, 0x00002a06, 0xffffd5fa, 0x0000062a, + 0xfffff9d6, 0x00002d2d, 0xffffd2d3, 0x000032fd, 0xffffcd03, 0xfffffd33, 0x000002cd, 0x000020e5, + 0xffffdf1b, 0xffffe521, 0x00001adf, 0x00002ceb, 0xffffd315, 0xffffeb2d, 0x000014d3, 0x00004521, + 0xffffbadf, 0x00002145, 0xffffdebb, 0x00004512, 0xffffbaee, 0x00001245, 0xffffedbb, 0x00004836, + 0xffffb7ca, 0x00003648, 0xffffc9b8, 0x00003eee, 0xffffc112, 0xffffee3f, 0x000011c1, 0x00004e06, + 0xffffb1fa, 0x0000064e, 0xfffff9b2, 0x00005151, 0xffffaeaf, 0x000032cd, 0xffffcd33, 0x00003ed6, + 0xffffc12a, 0xffffd63f, 0x000029c1, 0x000059f7, 0xffffa609, 0xfffff75a, 0x000008a6, 0x0000722a, + 0xffff8dd6, 0x00002a72, 0xffffd58e, 0x0000753f, 0xffff8ac1, 0x00003f75, 0xffffc08b, 0x000050dc, + 0xffffaf24, 0xffffdc51, 0x000023af, 0x00007815, 0xffff87eb, 0x00001578, 0xffffea88, 0x00007b60, + 0xffff84a0, 0x0000607b, 0xffff9f85, 0x00006ee2, 0xffff911e, 0xffffe26f, 0x00001d91, 0x00005cb2, + 0xffffa34e, 0xffffb25d, 0x00004da3, 0x000071bb, 0xffff8e45, 0xffffbb72, 0x0000448e, 0x00001212, + 0xffffedee, 0x00002121, 0xffffdedf, 0x00003f3f, 0xffffc0c1, 0x00006c6c, 0xffff9394, 0x00000000, + 0x03030000, 0xfcfd0000, 0x03000000, 0xfd000000, 0x00030000, 0xfffd0000, 0x06060000, 0xf9fa0000, + 0x00000303, 0x03030303, 0xfcfd0303, 0x03000303, 0xfd000303, 0x00030303, 0xfffd0303, 0x06060303, + 0xf9fa0303, 0xfffffcfd, 0x0302fcfd, 0xfcfcfcfd, 0x02fffcfd, 0xfcfffcfd, 0x0002fcfd, 0xfffcfcfd, + 0x0605fcfd, 0xf9f9fcfd, 0x00000300, 0x03030300, 0xfcfd0300, 0x03000300, 0xfd000300, 0x00030300, + 0xfffd0300, 0x06060300, 0xf9fa0300, 0xfffffd00, 0x0302fd00, 0xfcfcfd00, 0x02fffd00, 0xfcfffd00, + 0x0002fd00, 0xfffcfd00, 0x0605fd00, 0xf9f9fd00, 0x00000003, 0x03030003, 0xfcfd0003, 0x03000003, + 0xfd000003, 0x00030003, 0xfffd0003, 0x06060003, 0xf9fa0003, 0xfffffffd, 0x0302fffd, 0xfcfcfffd, + 0x02fffffd, 0xfcfffffd, 0x0002fffd, 0xfffcfffd, 0x0605fffd, 0xf9f9fffd, 0x00000606, 0x03030606, + 0xfcfd0606, 0x03000606, 0xfd000606, 0x00030606, 0xfffd0606, 0x06060606, 0xf9fa0606, 0xfffff9fa, + 0x0302f9fa, 0xfcfcf9fa, 0x02fff9fa, 0xfcfff9fa, 0x0002f9fa, 0xfffcf9fa, 0x0605f9fa, 0xf9f9f9fa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000404, 0xfffffbfc, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000804, + 0xfffff7fc, 0x00000408, 0xfffffbf8, 0x00000808, 0xfffff7f8, 0x000007f8, 0xfffff808, 0x00000bfc, + 0xfffff404, 0xfffffc0c, 0x000003f4, 0x00001008, 0xffffeff8, 0x00000810, 0xfffff7f0, 0x00001010, + 0xffffeff0, 0x00001400, 0xffffec00, 0x00000014, 0xffffffec, 0x00000ff4, 0xfffff00c, 0xfffff410, + 0x00000bf0, 0x000017fc, 0xffffe804, 0xfffffc18, 0x000003e8, 0x00002010, 0xffffdff0, 0x00001020, + 0xffffefe0, 0x00002008, 0xffffdff8, 0x00000820, 0xfffff7e0, 0x00002020, 0xffffdfe0, 0x00002400, + 0xffffdc00, 0x00000024, 0xffffffdc, 0x000017e8, 0xffffe818, 0x00001ff0, 0xffffe010, 0xfffff020, + 0x00000fe0, 0x00003414, 0xffffcbec, 0x00001434, 0xffffebcc, 0x00003824, 0xffffc7dc, 0x00002438, + 0xffffdbc8, 0x00002bf4, 0xffffd40c, 0xfffff42c, 0x00000bd4, 0x00003808, 0xffffc7f8, 0x00000838, + 0xfffff7c8, 0x00003c3c, 0xffffc3c4, 0x00003ffc, 0xffffc004, 0xfffffc40, 0x000003c0, 0x00002bdc, + 0xffffd424, 0xffffdc2c, 0x000023d4, 0x00003be4, 0xffffc41c, 0xffffe43c, 0x00001bc4, 0x00005c2c, + 0xffffa3d4, 0x00002c5c, 0xffffd3a4, 0x00005c18, 0xffffa3e8, 0x0000185c, 0xffffe7a4, 0x00006048, + 0xffff9fb8, 0x00004860, 0xffffb7a0, 0x000053ec, 0xffffac14, 0xffffec54, 0x000013ac, 0x00006408, + 0xffff9bf8, 0x00000864, 0xfffff79c, 0x00006c6c, 0xffff9394, 0x000043bc, 0xffffbc44, 0x000053c8, + 0xffffac38, 0xffffc854, 0x000037ac, 0x000077f4, 0xffff880c, 0xfffff478, 0x00000b88, 0x00006bd0, + 0xffff9430, 0xffffd06c, 0x00002f94, 0x00007b98, 0xffff8468, 0xffff987c, 0x00006784, 0x00001818, + 0xffffe7e8, 0x00002c2c, 0xffffd3d4, 0x00005454, 0xffffabac, 0x00000000, 0x04040000, 0xfbfc0000, + 0x04000000, 0xfc000000, 0x00040000, 0xfffc0000, 0x08040000, 0xf7fc0000, 0x04080000, 0x00000404, + 0x04040404, 0xfbfc0404, 0x04000404, 0xfc000404, 0x00040404, 0xfffc0404, 0x08040404, 0xf7fc0404, + 0x04080404, 0xfffffbfc, 0x0403fbfc, 0xfbfbfbfc, 0x03fffbfc, 0xfbfffbfc, 0x0003fbfc, 0xfffbfbfc, + 0x0803fbfc, 0xf7fbfbfc, 0x0407fbfc, 0x00000400, 0x04040400, 0xfbfc0400, 0x04000400, 0xfc000400, + 0x00040400, 0xfffc0400, 0x08040400, 0xf7fc0400, 0x04080400, 0xfffffc00, 0x0403fc00, 0xfbfbfc00, + 0x03fffc00, 0xfbfffc00, 0x0003fc00, 0xfffbfc00, 0x0803fc00, 0xf7fbfc00, 0x0407fc00, 0x00000004, + 0x04040004, 0xfbfc0004, 0x04000004, 0xfc000004, 0x00040004, 0xfffc0004, 0x08040004, 0xf7fc0004, + 0x04080004, 0xfffffffc, 0x0403fffc, 0xfbfbfffc, 0x03fffffc, 0xfbfffffc, 0x0003fffc, 0xfffbfffc, + 0x0803fffc, 0xf7fbfffc, 0x0407fffc, 0x00000804, 0x04040804, 0xfbfc0804, 0x04000804, 0xfc000804, + 0x00040804, 0xfffc0804, 0x08040804, 0xf7fc0804, 0x04080804, 0xfffff7fc, 0x0403f7fc, 0xfbfbf7fc, + 0x03fff7fc, 0xfbfff7fc, 0x0003f7fc, 0xfffbf7fc, 0x0803f7fc, 0xf7fbf7fc, 0x0407f7fc, 0x00000408, + 0x04040408, 0xfbfc0408, 0x04000408, 0xfc000408, 0x00040408, 0xfffc0408, 0x08040408, 0xf7fc0408, + 0x04080408, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000505, 0xfffffafb, 0x00000500, 0xfffffb00, 0x00000005, 0xfffffffb, 0x00000a0a, + 0xfffff5f6, 0x00000f05, 0xfffff0fb, 0x0000050f, 0xfffffaf1, 0x000009f6, 0xfffff60a, 0x00000efb, + 0xfffff105, 0xfffffb0f, 0x000004f1, 0x0000140a, 0xffffebf6, 0x00000a14, 0xfffff5ec, 0x00001414, + 0xffffebec, 0x00001900, 0xffffe700, 0x00000019, 0xffffffe7, 0x000013f1, 0xffffec0f, 0xfffff114, + 0x00000eec, 0x00002819, 0xffffd7e7, 0x00001928, 0xffffe6d8, 0x00001df6, 0xffffe20a, 0xfffff61e, + 0x000009e2, 0x0000280a, 0xffffd7f6, 0x00000a28, 0xfffff5d8, 0x00002828, 0xffffd7d8, 0x00002d00, + 0xffffd300, 0x0000002d, 0xffffffd3, 0x00001de2, 0xffffe21e, 0x000027ec, 0xffffd814, 0xffffec28, + 0x000013d8, 0x00004119, 0xffffbee7, 0x00001941, 0xffffe6bf, 0x0000462d, 0xffffb9d3, 0x00002d46, + 0xffffd2ba, 0x000036f1, 0xffffc90f, 0xfffff137, 0x00000ec9, 0x0000460a, 0xffffb9f6, 0x00000a46, + 0xfffff5ba, 0x00004b4b, 0xffffb4b5, 0x000054fb, 0xffffab05, 0xfffffb55, 0x000004ab, 0x000036d3, + 0xffffc92d, 0xffffd337, 0x00002cc9, 0x00004add, 0xffffb523, 0xffffdd4b, 0x000022b5, 0x00007337, + 0xffff8cc9, 0x00003773, 0xffffc88d, 0x0000731e, 0xffff8ce2, 0x00001e73, 0xffffe18d, 0x0000785a, + 0xffff87a6, 0x00005a78, 0xffffa588, 0x000068e2, 0xffff971e, 0xffffe269, 0x00001d97, 0x000054ab, + 0xffffab55, 0x000068ba, 0xffff9746, 0xffffba69, 0x00004597, 0x00001e1e, 0xffffe1e2, 0x00003c3c, + 0xffffc3c4, 0x00006969, 0xffff9697, 0x00000000, 0x05050000, 0xfafb0000, 0x05000000, 0xfb000000, + 0x00050000, 0xfffb0000, 0x0a0a0000, 0xf5f60000, 0x0f050000, 0xf0fb0000, 0x00000505, 0x05050505, + 0xfafb0505, 0x05000505, 0xfb000505, 0x00050505, 0xfffb0505, 0x0a0a0505, 0xf5f60505, 0x0f050505, + 0xf0fb0505, 0xfffffafb, 0x0504fafb, 0xfafafafb, 0x04fffafb, 0xfafffafb, 0x0004fafb, 0xfffafafb, + 0x0a09fafb, 0xf5f5fafb, 0x0f04fafb, 0xf0fafafb, 0x00000500, 0x05050500, 0xfafb0500, 0x05000500, + 0xfb000500, 0x00050500, 0xfffb0500, 0x0a0a0500, 0xf5f60500, 0x0f050500, 0xf0fb0500, 0xfffffb00, + 0x0504fb00, 0xfafafb00, 0x04fffb00, 0xfafffb00, 0x0004fb00, 0xfffafb00, 0x0a09fb00, 0xf5f5fb00, + 0x0f04fb00, 0xf0fafb00, 0x00000005, 0x05050005, 0xfafb0005, 0x05000005, 0xfb000005, 0x00050005, + 0xfffb0005, 0x0a0a0005, 0xf5f60005, 0x0f050005, 0xf0fb0005, 0xfffffffb, 0x0504fffb, 0xfafafffb, + 0x04fffffb, 0xfafffffb, 0x0004fffb, 0xfffafffb, 0x0a09fffb, 0xf5f5fffb, 0x0f04fffb, 0xf0fafffb, + 0x00000a0a, 0x05050a0a, 0xfafb0a0a, 0x05000a0a, 0xfb000a0a, 0x00050a0a, 0xfffb0a0a, 0x0a0a0a0a, + 0xf5f60a0a, 0x0f050a0a, 0xf0fb0a0a, 0xfffff5f6, 0x0504f5f6, 0xfafaf5f6, 0x04fff5f6, 0xfafff5f6, + 0x0004f5f6, 0xfffaf5f6, 0x0a09f5f6, 0xf5f5f5f6, 0x0f04f5f6, 0xf0faf5f6, 0x00000f05, 0x05050f05, + 0xfafb0f05, 0x05000f05, 0xfb000f05, 0x00050f05, 0xfffb0f05, 0x0a0a0f05, 0xf5f60f05, 0x0f050f05, + 0xf0fb0f05, 0xfffff0fb, 0x0504f0fb, 0xfafaf0fb, 0x04fff0fb, 0xfafff0fb, 0x0004f0fb, 0xfffaf0fb, + 0x0a09f0fb, 0xf5f5f0fb, 0x0f04f0fb, 0xf0faf0fb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000606, 0xfffff9fa, 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x00000c0c, + 0xfffff3f4, 0x00000c06, 0xfffff3fa, 0x0000060c, 0xfffff9f4, 0x00000bf4, 0xfffff40c, 0x000011fa, + 0xffffee06, 0xfffffa12, 0x000005ee, 0x0000180c, 0xffffe7f4, 0x00000c18, 0xfffff3e8, 0x00001818, + 0xffffe7e8, 0x00001e00, 0xffffe200, 0x0000001e, 0xffffffe2, 0x000017ee, 0xffffe812, 0xffffee18, + 0x000011e8, 0x0000301e, 0xffffcfe2, 0x00001e30, 0xffffe1d0, 0x000023fa, 0xffffdc06, 0xfffffa24, + 0x000005dc, 0x0000300c, 0xffffcff4, 0x00000c30, 0xfffff3d0, 0x00003030, 0xffffcfd0, 0x00003600, + 0xffffca00, 0x00000036, 0xffffffca, 0x000023dc, 0xffffdc24, 0x00002fe8, 0xffffd018, 0xffffe830, + 0x000017d0, 0x00004e1e, 0xffffb1e2, 0x00001e4e, 0xffffe1b2, 0x00005436, 0xffffabca, 0x00003654, + 0xffffc9ac, 0x000041ee, 0xffffbe12, 0xffffee42, 0x000011be, 0x0000540c, 0xffffabf4, 0x00000c54, + 0xfffff3ac, 0x00005a5a, 0xffffa5a6, 0x00005ffa, 0xffffa006, 0xfffffa60, 0x000005a0, 0x000041ca, + 0xffffbe36, 0xffffca42, 0x000035be, 0x000059d6, 0xffffa62a, 0xffffd65a, 0x000029a6, 0x00007de2, + 0xffff821e, 0xffffe27e, 0x00001d82, 0x0000659a, 0xffff9a66, 0x00007dac, 0xffff8254, 0xffffac7e, + 0x00005382, 0x00002424, 0xffffdbdc, 0x00004242, 0xffffbdbe, 0x00000000, 0x06060000, 0xf9fa0000, + 0x06000000, 0xfa000000, 0x00060000, 0xfffa0000, 0x0c0c0000, 0xf3f40000, 0x0c060000, 0xf3fa0000, + 0x060c0000, 0x00000606, 0x06060606, 0xf9fa0606, 0x06000606, 0xfa000606, 0x00060606, 0xfffa0606, + 0x0c0c0606, 0xf3f40606, 0x0c060606, 0xf3fa0606, 0x060c0606, 0xfffff9fa, 0x0605f9fa, 0xf9f9f9fa, + 0x05fff9fa, 0xf9fff9fa, 0x0005f9fa, 0xfff9f9fa, 0x0c0bf9fa, 0xf3f3f9fa, 0x0c05f9fa, 0xf3f9f9fa, + 0x060bf9fa, 0x00000600, 0x06060600, 0xf9fa0600, 0x06000600, 0xfa000600, 0x00060600, 0xfffa0600, + 0x0c0c0600, 0xf3f40600, 0x0c060600, 0xf3fa0600, 0x060c0600, 0xfffffa00, 0x0605fa00, 0xf9f9fa00, + 0x05fffa00, 0xf9fffa00, 0x0005fa00, 0xfff9fa00, 0x0c0bfa00, 0xf3f3fa00, 0x0c05fa00, 0xf3f9fa00, + 0x060bfa00, 0x00000006, 0x06060006, 0xf9fa0006, 0x06000006, 0xfa000006, 0x00060006, 0xfffa0006, + 0x0c0c0006, 0xf3f40006, 0x0c060006, 0xf3fa0006, 0x060c0006, 0xfffffffa, 0x0605fffa, 0xf9f9fffa, + 0x05fffffa, 0xf9fffffa, 0x0005fffa, 0xfff9fffa, 0x0c0bfffa, 0xf3f3fffa, 0x0c05fffa, 0xf3f9fffa, + 0x060bfffa, 0x00000c0c, 0x06060c0c, 0xf9fa0c0c, 0x06000c0c, 0xfa000c0c, 0x00060c0c, 0xfffa0c0c, + 0x0c0c0c0c, 0xf3f40c0c, 0x0c060c0c, 0xf3fa0c0c, 0x060c0c0c, 0xfffff3f4, 0x0605f3f4, 0xf9f9f3f4, + 0x05fff3f4, 0xf9fff3f4, 0x0005f3f4, 0xfff9f3f4, 0x0c0bf3f4, 0xf3f3f3f4, 0x0c05f3f4, 0xf3f9f3f4, + 0x060bf3f4, 0x00000c06, 0x06060c06, 0xf9fa0c06, 0x06000c06, 0xfa000c06, 0x00060c06, 0xfffa0c06, + 0x0c0c0c06, 0xf3f40c06, 0x0c060c06, 0xf3fa0c06, 0x060c0c06, 0xfffff3fa, 0x0605f3fa, 0xf9f9f3fa, + 0x05fff3fa, 0xf9fff3fa, 0x0005f3fa, 0xfff9f3fa, 0x0c0bf3fa, 0xf3f3f3fa, 0x0c05f3fa, 0xf3f9f3fa, + 0x060bf3fa, 0x0000060c, 0x0606060c, 0xf9fa060c, 0x0600060c, 0xfa00060c, 0x0006060c, 0xfffa060c, + 0x0c0c060c, 0xf3f4060c, 0x0c06060c, 0xf3fa060c, 0x060c060c, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000707, 0xfffff8f9, 0x00000700, 0xfffff900, 0x00000007, 0xfffffff9, 0x00000e0e, + 0xfffff1f2, 0x00001507, 0xffffeaf9, 0x00000715, 0xfffff8eb, 0x00000df2, 0xfffff20e, 0x000014f9, + 0xffffeb07, 0xfffff915, 0x000006eb, 0x00001c0e, 0xffffe3f2, 0x00000e1c, 0xfffff1e4, 0x00001c1c, + 0xffffe3e4, 0x00002300, 0xffffdd00, 0x00000023, 0xffffffdd, 0x00001beb, 0xffffe415, 0xffffeb1c, + 0x000014e4, 0x00003823, 0xffffc7dd, 0x00002338, 0xffffdcc8, 0x000029f2, 0xffffd60e, 0xfffff22a, + 0x00000dd6, 0x0000380e, 0xffffc7f2, 0x00000e38, 0xfffff1c8, 0x00003838, 0xffffc7c8, 0x00003f00, + 0xffffc100, 0x0000003f, 0xffffffc1, 0x000029d6, 0xffffd62a, 0x000037e4, 0xffffc81c, 0xffffe438, + 0x00001bc8, 0x00005b23, 0xffffa4dd, 0x0000235b, 0xffffdca5, 0x0000623f, 0xffff9dc1, 0x00003f62, + 0xffffc09e, 0x00004ceb, 0xffffb315, 0xffffeb4d, 0x000014b3, 0x0000620e, 0xffff9df2, 0x00000e62, + 0xfffff19e, 0x00006969, 0xffff9697, 0x000076f9, 0xffff8907, 0xfffff977, 0x00000689, 0x00004cc1, + 0xffffb33f, 0xffffc14d, 0x00003eb3, 0x000068cf, 0xffff9731, 0xffffcf69, 0x00003097, 0x00007689, + 0xffff8977, 0x00002a2a, 0xffffd5d6, 0x00004d4d, 0xffffb2b3, 0x00000000, 0x07070000, 0xf8f90000, + 0x07000000, 0xf9000000, 0x00070000, 0xfff90000, 0x0e0e0000, 0xf1f20000, 0x15070000, 0xeaf90000, + 0x07150000, 0x00000707, 0x07070707, 0xf8f90707, 0x07000707, 0xf9000707, 0x00070707, 0xfff90707, + 0x0e0e0707, 0xf1f20707, 0x15070707, 0xeaf90707, 0x07150707, 0xfffff8f9, 0x0706f8f9, 0xf8f8f8f9, + 0x06fff8f9, 0xf8fff8f9, 0x0006f8f9, 0xfff8f8f9, 0x0e0df8f9, 0xf1f1f8f9, 0x1506f8f9, 0xeaf8f8f9, + 0x0714f8f9, 0x00000700, 0x07070700, 0xf8f90700, 0x07000700, 0xf9000700, 0x00070700, 0xfff90700, + 0x0e0e0700, 0xf1f20700, 0x15070700, 0xeaf90700, 0x07150700, 0xfffff900, 0x0706f900, 0xf8f8f900, + 0x06fff900, 0xf8fff900, 0x0006f900, 0xfff8f900, 0x0e0df900, 0xf1f1f900, 0x1506f900, 0xeaf8f900, + 0x0714f900, 0x00000007, 0x07070007, 0xf8f90007, 0x07000007, 0xf9000007, 0x00070007, 0xfff90007, + 0x0e0e0007, 0xf1f20007, 0x15070007, 0xeaf90007, 0x07150007, 0xfffffff9, 0x0706fff9, 0xf8f8fff9, + 0x06fffff9, 0xf8fffff9, 0x0006fff9, 0xfff8fff9, 0x0e0dfff9, 0xf1f1fff9, 0x1506fff9, 0xeaf8fff9, + 0x0714fff9, 0x00000e0e, 0x07070e0e, 0xf8f90e0e, 0x07000e0e, 0xf9000e0e, 0x00070e0e, 0xfff90e0e, + 0x0e0e0e0e, 0xf1f20e0e, 0x15070e0e, 0xeaf90e0e, 0x07150e0e, 0xfffff1f2, 0x0706f1f2, 0xf8f8f1f2, + 0x06fff1f2, 0xf8fff1f2, 0x0006f1f2, 0xfff8f1f2, 0x0e0df1f2, 0xf1f1f1f2, 0x1506f1f2, 0xeaf8f1f2, + 0x0714f1f2, 0x00001507, 0x07071507, 0xf8f91507, 0x07001507, 0xf9001507, 0x00071507, 0xfff91507, + 0x0e0e1507, 0xf1f21507, 0x15071507, 0xeaf91507, 0x07151507, 0xffffeaf9, 0x0706eaf9, 0xf8f8eaf9, + 0x06ffeaf9, 0xf8ffeaf9, 0x0006eaf9, 0xfff8eaf9, 0x0e0deaf9, 0xf1f1eaf9, 0x1506eaf9, 0xeaf8eaf9, + 0x0714eaf9, 0x00000715, 0x07070715, 0xf8f90715, 0x07000715, 0xf9000715, 0x00070715, 0xfff90715, + 0x0e0e0715, 0xf1f20715, 0x15070715, 0xeaf90715, 0x07150715, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000808, 0xfffff7f8, 0x00000800, 0xfffff800, 0x00000008, 0xfffffff8, 0x00001010, + 0xffffeff0, 0x00001008, 0xffffeff8, 0x00000810, 0xfffff7f0, 0x00000ff0, 0xfffff010, 0x000017f8, + 0xffffe808, 0xfffff818, 0x000007e8, 0x00002010, 0xffffdff0, 0x00001020, 0xffffefe0, 0x00002020, + 0xffffdfe0, 0x00002800, 0xffffd800, 0x00000028, 0xffffffd8, 0x00001fe8, 0xffffe018, 0xffffe820, + 0x000017e0, 0x00004028, 0xffffbfd8, 0x00002840, 0xffffd7c0, 0x00002ff0, 0xffffd010, 0xfffff030, + 0x00000fd0, 0x00004010, 0xffffbff0, 0x00001040, 0xffffefc0, 0x00004040, 0xffffbfc0, 0x00004800, + 0xffffb800, 0x00000048, 0xffffffb8, 0x00002fd0, 0xffffd030, 0x00003fe0, 0xffffc020, 0xffffe040, + 0x00001fc0, 0x00006828, 0xffff97d8, 0x00002868, 0xffffd798, 0x00007048, 0xffff8fb8, 0x00004870, + 0xffffb790, 0x000057e8, 0xffffa818, 0xffffe858, 0x000017a8, 0x00007010, 0xffff8ff0, 0x00001070, + 0xffffef90, 0x00007878, 0xffff8788, 0x000057b8, 0xffffa848, 0xffffb858, 0x000047a8, 0x000077c8, + 0xffff8838, 0xffffc878, 0x00003788, 0x00003030, 0xffffcfd0, 0x00005858, 0xffffa7a8, 0x00000000, + 0x08080000, 0xf7f80000, 0x08000000, 0xf8000000, 0x00080000, 0xfff80000, 0x10100000, 0xeff00000, + 0x10080000, 0xeff80000, 0x08100000, 0x00000808, 0x08080808, 0xf7f80808, 0x08000808, 0xf8000808, + 0x00080808, 0xfff80808, 0x10100808, 0xeff00808, 0x10080808, 0xeff80808, 0x08100808, 0xfffff7f8, + 0x0807f7f8, 0xf7f7f7f8, 0x07fff7f8, 0xf7fff7f8, 0x0007f7f8, 0xfff7f7f8, 0x100ff7f8, 0xefeff7f8, + 0x1007f7f8, 0xeff7f7f8, 0x080ff7f8, 0x00000800, 0x08080800, 0xf7f80800, 0x08000800, 0xf8000800, + 0x00080800, 0xfff80800, 0x10100800, 0xeff00800, 0x10080800, 0xeff80800, 0x08100800, 0xfffff800, + 0x0807f800, 0xf7f7f800, 0x07fff800, 0xf7fff800, 0x0007f800, 0xfff7f800, 0x100ff800, 0xefeff800, + 0x1007f800, 0xeff7f800, 0x080ff800, 0x00000008, 0x08080008, 0xf7f80008, 0x08000008, 0xf8000008, + 0x00080008, 0xfff80008, 0x10100008, 0xeff00008, 0x10080008, 0xeff80008, 0x08100008, 0xfffffff8, + 0x0807fff8, 0xf7f7fff8, 0x07fffff8, 0xf7fffff8, 0x0007fff8, 0xfff7fff8, 0x100ffff8, 0xefeffff8, + 0x1007fff8, 0xeff7fff8, 0x080ffff8, 0x00001010, 0x08081010, 0xf7f81010, 0x08001010, 0xf8001010, + 0x00081010, 0xfff81010, 0x10101010, 0xeff01010, 0x10081010, 0xeff81010, 0x08101010, 0xffffeff0, + 0x0807eff0, 0xf7f7eff0, 0x07ffeff0, 0xf7ffeff0, 0x0007eff0, 0xfff7eff0, 0x100feff0, 0xefefeff0, + 0x1007eff0, 0xeff7eff0, 0x080feff0, 0x00001008, 0x08081008, 0xf7f81008, 0x08001008, 0xf8001008, + 0x00081008, 0xfff81008, 0x10101008, 0xeff01008, 0x10081008, 0xeff81008, 0x08101008, 0xffffeff8, + 0x0807eff8, 0xf7f7eff8, 0x07ffeff8, 0xf7ffeff8, 0x0007eff8, 0xfff7eff8, 0x100feff8, 0xefefeff8, + 0x1007eff8, 0xeff7eff8, 0x080feff8, 0x00000810, 0x08080810, 0xf7f80810, 0x08000810, 0xf8000810, + 0x00080810, 0xfff80810, 0x10100810, 0xeff00810, 0x10080810, 0xeff80810, 0x08100810, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000909, 0xfffff6f7, 0x00000900, 0xfffff700, 0x00000009, 0xfffffff7, 0x00001212, + 0xffffedee, 0x00001b09, 0xffffe4f7, 0x0000091b, 0xfffff6e5, 0x000011ee, 0xffffee12, 0x00001af7, + 0xffffe509, 0xfffff71b, 0x000008e5, 0x00002412, 0xffffdbee, 0x00001224, 0xffffeddc, 0x00002424, + 0xffffdbdc, 0x00002d00, 0xffffd300, 0x0000002d, 0xffffffd3, 0x000023e5, 0xffffdc1b, 0xffffe524, + 0x00001adc, 0x0000482d, 0xffffb7d3, 0x00002d48, 0xffffd2b8, 0x000035ee, 0xffffca12, 0xffffee36, + 0x000011ca, 0x00004812, 0xffffb7ee, 0x00001248, 0xffffedb8, 0x00004848, 0xffffb7b8, 0x00005100, + 0xffffaf00, 0x00000051, 0xffffffaf, 0x000035ca, 0xffffca36, 0x000047dc, 0xffffb824, 0xffffdc48, + 0x000023b8, 0x0000752d, 0xffff8ad3, 0x00002d75, 0xffffd28b, 0x00007e51, 0xffff81af, 0x0000517e, + 0xffffae82, 0x000062e5, 0xffff9d1b, 0xffffe563, 0x00001a9d, 0x000062af, 0xffff9d51, 0xffffaf63, + 0x0000509d, 0x00003636, 0xffffc9ca, 0x00006c6c, 0xffff9394, 0x00000000, 0x09090000, 0xf6f70000, + 0x09000000, 0xf7000000, 0x00090000, 0xfff70000, 0x12120000, 0xedee0000, 0x1b090000, 0xe4f70000, + 0x091b0000, 0xf6e50000, 0x00000909, 0x09090909, 0xf6f70909, 0x09000909, 0xf7000909, 0x00090909, + 0xfff70909, 0x12120909, 0xedee0909, 0x1b090909, 0xe4f70909, 0x091b0909, 0xf6e50909, 0xfffff6f7, + 0x0908f6f7, 0xf6f6f6f7, 0x08fff6f7, 0xf6fff6f7, 0x0008f6f7, 0xfff6f6f7, 0x1211f6f7, 0xededf6f7, + 0x1b08f6f7, 0xe4f6f6f7, 0x091af6f7, 0xf6e4f6f7, 0x00000900, 0x09090900, 0xf6f70900, 0x09000900, + 0xf7000900, 0x00090900, 0xfff70900, 0x12120900, 0xedee0900, 0x1b090900, 0xe4f70900, 0x091b0900, + 0xf6e50900, 0xfffff700, 0x0908f700, 0xf6f6f700, 0x08fff700, 0xf6fff700, 0x0008f700, 0xfff6f700, + 0x1211f700, 0xededf700, 0x1b08f700, 0xe4f6f700, 0x091af700, 0xf6e4f700, 0x00000009, 0x09090009, + 0xf6f70009, 0x09000009, 0xf7000009, 0x00090009, 0xfff70009, 0x12120009, 0xedee0009, 0x1b090009, + 0xe4f70009, 0x091b0009, 0xf6e50009, 0xfffffff7, 0x0908fff7, 0xf6f6fff7, 0x08fffff7, 0xf6fffff7, + 0x0008fff7, 0xfff6fff7, 0x1211fff7, 0xededfff7, 0x1b08fff7, 0xe4f6fff7, 0x091afff7, 0xf6e4fff7, + 0x00001212, 0x09091212, 0xf6f71212, 0x09001212, 0xf7001212, 0x00091212, 0xfff71212, 0x12121212, + 0xedee1212, 0x1b091212, 0xe4f71212, 0x091b1212, 0xf6e51212, 0xffffedee, 0x0908edee, 0xf6f6edee, + 0x08ffedee, 0xf6ffedee, 0x0008edee, 0xfff6edee, 0x1211edee, 0xedededee, 0x1b08edee, 0xe4f6edee, + 0x091aedee, 0xf6e4edee, 0x00001b09, 0x09091b09, 0xf6f71b09, 0x09001b09, 0xf7001b09, 0x00091b09, + 0xfff71b09, 0x12121b09, 0xedee1b09, 0x1b091b09, 0xe4f71b09, 0x091b1b09, 0xf6e51b09, 0xffffe4f7, + 0x0908e4f7, 0xf6f6e4f7, 0x08ffe4f7, 0xf6ffe4f7, 0x0008e4f7, 0xfff6e4f7, 0x1211e4f7, 0xedede4f7, + 0x1b08e4f7, 0xe4f6e4f7, 0x091ae4f7, 0xf6e4e4f7, 0x0000091b, 0x0909091b, 0xf6f7091b, 0x0900091b, + 0xf700091b, 0x0009091b, 0xfff7091b, 0x1212091b, 0xedee091b, 0x1b09091b, 0xe4f7091b, 0x091b091b, + 0xf6e5091b, 0xfffff6e5, 0x0908f6e5, 0xf6f6f6e5, 0x08fff6e5, 0xf6fff6e5, 0x0008f6e5, 0xfff6f6e5, + 0x1211f6e5, 0xededf6e5, 0x1b08f6e5, 0xe4f6f6e5, 0x091af6e5, 0xf6e4f6e5, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, + 0xfffff9fa, 0x00000700, 0xfffff900, 0x00000007, 0xfffffff9, 0x000004fb, 0xfffffb05, 0xfffffb05, + 0x000004fb, 0x00000b06, 0xfffff4fa, 0x0000060b, 0xfffff9f5, 0x00000800, 0xfffff800, 0x00000008, + 0xfffffff8, 0x00000b0b, 0xfffff4f5, 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x0000110c, + 0xffffeef4, 0x00000c11, 0xfffff3ef, 0x00001111, 0xffffeeef, 0x00001206, 0xffffedfa, 0x00000612, + 0xfffff9ee, 0x00000af8, 0xfffff508, 0xfffff80b, 0x000007f5, 0x00000f00, 0xfffff100, 0x0000000f, + 0xfffffff1, 0x00001400, 0xffffec00, 0x00000014, 0xffffffec, 0x00001912, 0xffffe6ee, 0x00001219, + 0xffffede7, 0x0000190b, 0xffffe6f5, 0x00000b19, 0xfffff4e7, 0x00001919, 0xffffe6e7, 0x00000df2, + 0xfffff20e, 0xfffff20e, 0x00000df2, 0x00001a00, 0xffffe600, 0x0000001a, 0xffffffe6, 0x000011f5, + 0xffffee0b, 0xfffff512, 0x00000aee, 0x000015f9, 0xffffea07, 0xfffff916, 0x000006ea, 0x0000221a, + 0xffffdde6, 0x00001a22, 0xffffe5de, 0x00002212, 0xffffddee, 0x00001222, 0xffffedde, 0x00002222, + 0xffffddde, 0x0000230b, 0xffffdcf5, 0x00000b23, 0xfffff4dd, 0x00001d00, 0xffffe300, 0x0000001d, + 0xffffffe3, 0x000015ed, 0xffffea13, 0xffffed16, 0x000012ea, 0x000019f1, 0xffffe60f, 0xfffff11a, + 0x00000ee6, 0x00002500, 0xffffdb00, 0x00000025, 0xffffffdb, 0x00002c1b, 0xffffd3e5, 0x00001b2c, + 0xffffe4d4, 0x00002c24, 0xffffd3dc, 0x0000242c, 0xffffdbd4, 0x00002c12, 0xffffd3ee, 0x0000122c, + 0xffffedd4, 0x000020f6, 0xffffdf0a, 0xfffff621, 0x000009df, 0x00002d2d, 0xffffd2d3, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000300, 0xfffffd00, 0x00000003, 0xfffffffd, 0x00000606, + 0xfffff9fa, 0x00000700, 0xfffff900, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020300, 0x0201fd00, + 0x02020003, 0x0201fffd, 0x02020606, 0x0201f9fa, 0x02020700, 0x0201f900, 0xfdfe0000, 0xfdfe0202, + 0xfdfdfdfe, 0xfdfe0300, 0xfdfdfd00, 0xfdfe0003, 0xfdfdfffd, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0700, + 0xfdfdf900, 0x03000000, 0x03000202, 0x02fffdfe, 0x03000300, 0x02fffd00, 0x03000003, 0x02fffffd, + 0x03000606, 0x02fff9fa, 0x03000700, 0x02fff900, 0xfd000000, 0xfd000202, 0xfcfffdfe, 0xfd000300, + 0xfcfffd00, 0xfd000003, 0xfcfffffd, 0xfd000606, 0xfcfff9fa, 0xfd000700, 0xfcfff900, 0x00030000, + 0x00030202, 0x0002fdfe, 0x00030300, 0x0002fd00, 0x00030003, 0x0002fffd, 0x00030606, 0x0002f9fa, + 0x00030700, 0x0002f900, 0xfffd0000, 0xfffd0202, 0xfffcfdfe, 0xfffd0300, 0xfffcfd00, 0xfffd0003, + 0xfffcfffd, 0xfffd0606, 0xfffcf9fa, 0xfffd0700, 0xfffcf900, 0x06060000, 0x06060202, 0x0605fdfe, + 0x06060300, 0x0605fd00, 0x06060003, 0x0605fffd, 0x06060606, 0x0605f9fa, 0x06060700, 0x0605f900, + 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0300, 0xf9f9fd00, 0xf9fa0003, 0xf9f9fffd, 0xf9fa0606, + 0xf9f9f9fa, 0xf9fa0700, 0xf9f9f900, 0x07000000, 0x07000202, 0x06fffdfe, 0x07000300, 0x06fffd00, + 0x07000003, 0x06fffffd, 0x07000606, 0x06fff9fa, 0x07000700, 0x06fff900, 0xf9000000, 0xf9000202, + 0xf8fffdfe, 0xf9000300, 0xf8fffd00, 0xf9000003, 0xf8fffffd, 0xf9000606, 0xf8fff9fa, 0xf9000700, + 0xf8fff900, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000202, 0xfffffdfe, 0x00000606, + 0xfffff9fa, 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x000003fc, 0xfffffc04, 0xfffffa0a, + 0x000005f6, 0xfffff400, 0x00000c00, 0xfffff3fa, 0xfffff406, 0x00000bfa, 0x00000c06, 0xfffffff2, + 0x0000000e, 0x00000c0c, 0xfffff3f4, 0xffffee00, 0x00001200, 0xfffff40e, 0x00000bf2, 0xfffff9ee, + 0xfffffa12, 0x000005ee, 0x00000612, 0xffffedf6, 0xffffee0a, 0x000011f6, 0x0000120a, 0xffffffea, + 0x00000016, 0xffffe800, 0x00001800, 0xfffff3ea, 0xfffff416, 0x00000bea, 0x00000c16, 0xffffe7f8, + 0xffffe808, 0x000017f8, 0x00001808, 0xfffff9e6, 0xfffffa1a, 0x000005e6, 0x0000061a, 0xffffffe4, + 0x0000001c, 0x00001414, 0xffffebec, 0xffffe5f2, 0x00001a0e, 0xfffff3e2, 0x00000c1e, 0xffffdff6, + 0x0000200a, 0xffffdfee, 0x00002012, 0xffffe5e6, 0x00001a1a, 0xffffebde, 0x00001422, 0xfffff3da, + 0x00000c26, 0xffffdfe0, 0x00002020, 0x00002020, 0xffffd7ea, 0xffffddde, 0x00002222, 0x00000000, + 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, + 0x00000600, 0xfffffa00, 0x00000006, 0xfffffffa, 0x02000000, 0x02000200, 0x01fffe00, 0x02000002, + 0x01fffffe, 0x02000202, 0x01fffdfe, 0x02000606, 0x01fff9fa, 0x02000600, 0x01fffa00, 0x02000006, + 0x01fffffa, 0xfe000000, 0xfe000200, 0xfdfffe00, 0xfe000002, 0xfdfffffe, 0xfe000202, 0xfdfffdfe, + 0xfe000606, 0xfdfff9fa, 0xfe000600, 0xfdfffa00, 0xfe000006, 0xfdfffffa, 0x00020000, 0x00020200, + 0x0001fe00, 0x00020002, 0x0001fffe, 0x00020202, 0x0001fdfe, 0x00020606, 0x0001f9fa, 0x00020600, + 0x0001fa00, 0x00020006, 0x0001fffa, 0xfffe0000, 0xfffe0200, 0xfffdfe00, 0xfffe0002, 0xfffdfffe, + 0xfffe0202, 0xfffdfdfe, 0xfffe0606, 0xfffdf9fa, 0xfffe0600, 0xfffdfa00, 0xfffe0006, 0xfffdfffa, + 0x02020000, 0x02020200, 0x0201fe00, 0x02020002, 0x0201fffe, 0x02020202, 0x0201fdfe, 0x02020606, + 0x0201f9fa, 0x02020600, 0x0201fa00, 0x02020006, 0x0201fffa, 0xfdfe0000, 0xfdfe0200, 0xfdfdfe00, + 0xfdfe0002, 0xfdfdfffe, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0600, 0xfdfdfa00, + 0xfdfe0006, 0xfdfdfffa, 0x06060000, 0x06060200, 0x0605fe00, 0x06060002, 0x0605fffe, 0x06060202, + 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060600, 0x0605fa00, 0x06060006, 0x0605fffa, 0xf9fa0000, + 0xf9fa0200, 0xf9f9fe00, 0xf9fa0002, 0xf9f9fffe, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, + 0xf9fa0600, 0xf9f9fa00, 0xf9fa0006, 0xf9f9fffa, 0x06000000, 0x06000200, 0x05fffe00, 0x06000002, + 0x05fffffe, 0x06000202, 0x05fffdfe, 0x06000606, 0x05fff9fa, 0x06000600, 0x05fffa00, 0x06000006, + 0x05fffffa, 0xfa000000, 0xfa000200, 0xf9fffe00, 0xfa000002, 0xf9fffffe, 0xfa000202, 0xf9fffdfe, + 0xfa000606, 0xf9fff9fa, 0xfa000600, 0xf9fffa00, 0xfa000006, 0xf9fffffa, 0x00060000, 0x00060200, + 0x0005fe00, 0x00060002, 0x0005fffe, 0x00060202, 0x0005fdfe, 0x00060606, 0x0005f9fa, 0x00060600, + 0x0005fa00, 0x00060006, 0x0005fffa, 0xfffa0000, 0xfffa0200, 0xfff9fe00, 0xfffa0002, 0xfff9fffe, + 0xfffa0202, 0xfff9fdfe, 0xfffa0606, 0xfff9f9fa, 0xfffa0600, 0xfff9fa00, 0xfffa0006, 0xfff9fffa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, 0xfffffbfc, 0x00000a0a, + 0xfffff5f6, 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x000005fa, 0xfffffa06, 0xfffff80e, + 0x000007f2, 0xffffffee, 0x00000012, 0xfffff00a, 0x00000ff6, 0xffffe800, 0x00001800, 0xfffff7e8, + 0xfffff818, 0x000007e8, 0x00000818, 0x00001212, 0xffffedee, 0xfffff014, 0x00000fec, 0xffffe5f2, + 0xffffe60e, 0x000019f2, 0x00001a0e, 0xffffffe2, 0x0000001e, 0xffffde00, 0x00002200, 0xfffff7de, + 0xfffff822, 0x000007de, 0x00000822, 0xffffede2, 0xffffee1e, 0x000011e2, 0x0000121e, 0xffffddf6, + 0xffffde0a, 0x000021f6, 0x0000220a, 0xffffddec, 0x00002214, 0xffffffd8, 0x00000028, 0x00001e1e, + 0xffffe1e2, 0xffffedd8, 0x00001228, 0xffffd400, 0x00002c00, 0xffffd3f0, 0x00002c10, 0xffffdbdc, + 0xffffdbdc, 0x00002424, 0xffffd3e6, 0x00002c1a, 0xffffe5d2, 0x00001a2e, 0xffffedcc, 0x00001234, + 0xffffc9ec, 0xffffd3d4, 0x00002c2c, 0xffffc9e0, 0xffffd1d2, 0xffffd1d2, 0x00002e2e, 0x00000000, + 0x00000200, 0xfffffe00, 0x00000002, 0xfffffffe, 0x00000404, 0xfffffbfc, 0x00000a0a, 0xfffff5f6, + 0x00000a00, 0xfffff600, 0x0000000a, 0xfffffff6, 0x02000000, 0x02000200, 0x01fffe00, 0x02000002, + 0x01fffffe, 0x02000404, 0x01fffbfc, 0x02000a0a, 0x01fff5f6, 0x02000a00, 0x01fff600, 0x0200000a, + 0x01fffff6, 0xfe000000, 0xfe000200, 0xfdfffe00, 0xfe000002, 0xfdfffffe, 0xfe000404, 0xfdfffbfc, + 0xfe000a0a, 0xfdfff5f6, 0xfe000a00, 0xfdfff600, 0xfe00000a, 0xfdfffff6, 0x00020000, 0x00020200, + 0x0001fe00, 0x00020002, 0x0001fffe, 0x00020404, 0x0001fbfc, 0x00020a0a, 0x0001f5f6, 0x00020a00, + 0x0001f600, 0x0002000a, 0x0001fff6, 0xfffe0000, 0xfffe0200, 0xfffdfe00, 0xfffe0002, 0xfffdfffe, + 0xfffe0404, 0xfffdfbfc, 0xfffe0a0a, 0xfffdf5f6, 0xfffe0a00, 0xfffdf600, 0xfffe000a, 0xfffdfff6, + 0x04040000, 0x04040200, 0x0403fe00, 0x04040002, 0x0403fffe, 0x04040404, 0x0403fbfc, 0x04040a0a, + 0x0403f5f6, 0x04040a00, 0x0403f600, 0x0404000a, 0x0403fff6, 0xfbfc0000, 0xfbfc0200, 0xfbfbfe00, + 0xfbfc0002, 0xfbfbfffe, 0xfbfc0404, 0xfbfbfbfc, 0xfbfc0a0a, 0xfbfbf5f6, 0xfbfc0a00, 0xfbfbf600, + 0xfbfc000a, 0xfbfbfff6, 0x0a0a0000, 0x0a0a0200, 0x0a09fe00, 0x0a0a0002, 0x0a09fffe, 0x0a0a0404, + 0x0a09fbfc, 0x0a0a0a0a, 0x0a09f5f6, 0x0a0a0a00, 0x0a09f600, 0x0a0a000a, 0x0a09fff6, 0xf5f60000, + 0xf5f60200, 0xf5f5fe00, 0xf5f60002, 0xf5f5fffe, 0xf5f60404, 0xf5f5fbfc, 0xf5f60a0a, 0xf5f5f5f6, + 0xf5f60a00, 0xf5f5f600, 0xf5f6000a, 0xf5f5fff6, 0x0a000000, 0x0a000200, 0x09fffe00, 0x0a000002, + 0x09fffffe, 0x0a000404, 0x09fffbfc, 0x0a000a0a, 0x09fff5f6, 0x0a000a00, 0x09fff600, 0x0a00000a, + 0x09fffff6, 0xf6000000, 0xf6000200, 0xf5fffe00, 0xf6000002, 0xf5fffffe, 0xf6000404, 0xf5fffbfc, + 0xf6000a0a, 0xf5fff5f6, 0xf6000a00, 0xf5fff600, 0xf600000a, 0xf5fffff6, 0x000a0000, 0x000a0200, + 0x0009fe00, 0x000a0002, 0x0009fffe, 0x000a0404, 0x0009fbfc, 0x000a0a0a, 0x0009f5f6, 0x000a0a00, + 0x0009f600, 0x000a000a, 0x0009fff6, 0xfff60000, 0xfff60200, 0xfff5fe00, 0xfff60002, 0xfff5fffe, + 0xfff60404, 0xfff5fbfc, 0xfff60a0a, 0xfff5f5f6, 0xfff60a00, 0xfff5f600, 0xfff6000a, 0xfff5fff6, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000404, 0xfffffbfc, 0x00000c0c, + 0xfffff3f4, 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x000007f8, 0xfffff808, 0xfffff008, + 0x00000ff8, 0xffffe800, 0x00001800, 0xfffff7e8, 0xfffff818, 0x000007e8, 0x00000818, 0xfffff014, + 0x00000fec, 0xffffffe4, 0x0000001c, 0xffffe7f0, 0xffffe810, 0x000017f0, 0x00001810, 0xffffe000, + 0x00002000, 0xffffefe4, 0xfffff01c, 0x00000fe4, 0x0000101c, 0xffffdff8, 0xffffe008, 0xfffff7e0, + 0xfffff820, 0x000007e0, 0x00000820, 0x00001ff8, 0x00002008, 0x00001818, 0xffffe7e8, 0xffffe818, + 0x000017e8, 0xffffdfec, 0x00002014, 0xffffffd8, 0x00000028, 0xffffefd8, 0x00001028, 0xffffd400, + 0xffffd400, 0xffffffd4, 0x0000002c, 0x00002c00, 0x00002c00, 0xffffdfe0, 0x00002020, 0xffffd3f0, + 0x00002c10, 0xffffd3e8, 0xffffe7d4, 0x0000182c, 0x00002c18, 0xffffefd0, 0x00001030, 0xffffdbdc, + 0xffffdbdc, 0x00002424, 0x00002424, 0xffffcbec, 0x00002828, 0xffffd7d8, 0xffffcbe0, 0x00000000, + 0x00000400, 0xfffffc00, 0x00000004, 0xfffffffc, 0x00000404, 0xfffffbfc, 0x00000c0c, 0xfffff3f4, + 0x00000c00, 0xfffff400, 0x0000000c, 0xfffffff4, 0x04000000, 0x04000400, 0x03fffc00, 0x04000004, + 0x03fffffc, 0x04000404, 0x03fffbfc, 0x04000c0c, 0x03fff3f4, 0x04000c00, 0x03fff400, 0x0400000c, + 0x03fffff4, 0xfc000000, 0xfc000400, 0xfbfffc00, 0xfc000004, 0xfbfffffc, 0xfc000404, 0xfbfffbfc, + 0xfc000c0c, 0xfbfff3f4, 0xfc000c00, 0xfbfff400, 0xfc00000c, 0xfbfffff4, 0x00040000, 0x00040400, + 0x0003fc00, 0x00040004, 0x0003fffc, 0x00040404, 0x0003fbfc, 0x00040c0c, 0x0003f3f4, 0x00040c00, + 0x0003f400, 0x0004000c, 0x0003fff4, 0xfffc0000, 0xfffc0400, 0xfffbfc00, 0xfffc0004, 0xfffbfffc, + 0xfffc0404, 0xfffbfbfc, 0xfffc0c0c, 0xfffbf3f4, 0xfffc0c00, 0xfffbf400, 0xfffc000c, 0xfffbfff4, + 0x04040000, 0x04040400, 0x0403fc00, 0x04040004, 0x0403fffc, 0x04040404, 0x0403fbfc, 0x04040c0c, + 0x0403f3f4, 0x04040c00, 0x0403f400, 0x0404000c, 0x0403fff4, 0xfbfc0000, 0xfbfc0400, 0xfbfbfc00, + 0xfbfc0004, 0xfbfbfffc, 0xfbfc0404, 0xfbfbfbfc, 0xfbfc0c0c, 0xfbfbf3f4, 0xfbfc0c00, 0xfbfbf400, + 0xfbfc000c, 0xfbfbfff4, 0x0c0c0000, 0x0c0c0400, 0x0c0bfc00, 0x0c0c0004, 0x0c0bfffc, 0x0c0c0404, + 0x0c0bfbfc, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c0c00, 0x0c0bf400, 0x0c0c000c, 0x0c0bfff4, 0xf3f40000, + 0xf3f40400, 0xf3f3fc00, 0xf3f40004, 0xf3f3fffc, 0xf3f40404, 0xf3f3fbfc, 0xf3f40c0c, 0xf3f3f3f4, + 0xf3f40c00, 0xf3f3f400, 0xf3f4000c, 0xf3f3fff4, 0x0c000000, 0x0c000400, 0x0bfffc00, 0x0c000004, + 0x0bfffffc, 0x0c000404, 0x0bfffbfc, 0x0c000c0c, 0x0bfff3f4, 0x0c000c00, 0x0bfff400, 0x0c00000c, + 0x0bfffff4, 0xf4000000, 0xf4000400, 0xf3fffc00, 0xf4000004, 0xf3fffffc, 0xf4000404, 0xf3fffbfc, + 0xf4000c0c, 0xf3fff3f4, 0xf4000c00, 0xf3fff400, 0xf400000c, 0xf3fffff4, 0x000c0000, 0x000c0400, + 0x000bfc00, 0x000c0004, 0x000bfffc, 0x000c0404, 0x000bfbfc, 0x000c0c0c, 0x000bf3f4, 0x000c0c00, + 0x000bf400, 0x000c000c, 0x000bfff4, 0xfff40000, 0xfff40400, 0xfff3fc00, 0xfff40004, 0xfff3fffc, + 0xfff40404, 0xfff3fbfc, 0xfff40c0c, 0xfff3f3f4, 0xfff40c00, 0xfff3f400, 0xfff4000c, 0xfff3fff4, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, + 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, + 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, + 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, + 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, + 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, + 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, + 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, + 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, + 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, + 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, + 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, + 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, + 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, + 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, + 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, + 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, + 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, + 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, + 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, + 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, + 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, + 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, + 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, + 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, + 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, + 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, + 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, + 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, + 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, + 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, + 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, + 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, + 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, + 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, + 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, + 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, + 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, + 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, + 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, + 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, + 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, + 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, + 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, + 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, + 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, + 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, + 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, + 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, + 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, + 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, + 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, + 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, + 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, + 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, + 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, + 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, + 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, + 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, + 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, + 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, + 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, + 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, + 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, + 0xffffebec, 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000202, 0xfffffdfe, 0x00000606, 0xfffff9fa, 0x00000c0c, 0xfffff3f4, 0x00001414, 0xffffebec, + 0x00002020, 0xffffdfe0, 0x00002e2e, 0xffffd1d2, 0x02020000, 0x02020202, 0x0201fdfe, 0x02020606, + 0x0201f9fa, 0x02020c0c, 0x0201f3f4, 0x02021414, 0x0201ebec, 0x02022020, 0x0201dfe0, 0x02022e2e, + 0x0201d1d2, 0xfdfe0000, 0xfdfe0202, 0xfdfdfdfe, 0xfdfe0606, 0xfdfdf9fa, 0xfdfe0c0c, 0xfdfdf3f4, + 0xfdfe1414, 0xfdfdebec, 0xfdfe2020, 0xfdfddfe0, 0xfdfe2e2e, 0xfdfdd1d2, 0x06060000, 0x06060202, + 0x0605fdfe, 0x06060606, 0x0605f9fa, 0x06060c0c, 0x0605f3f4, 0x06061414, 0x0605ebec, 0x06062020, + 0x0605dfe0, 0x06062e2e, 0x0605d1d2, 0xf9fa0000, 0xf9fa0202, 0xf9f9fdfe, 0xf9fa0606, 0xf9f9f9fa, + 0xf9fa0c0c, 0xf9f9f3f4, 0xf9fa1414, 0xf9f9ebec, 0xf9fa2020, 0xf9f9dfe0, 0xf9fa2e2e, 0xf9f9d1d2, + 0x0c0c0000, 0x0c0c0202, 0x0c0bfdfe, 0x0c0c0606, 0x0c0bf9fa, 0x0c0c0c0c, 0x0c0bf3f4, 0x0c0c1414, + 0x0c0bebec, 0x0c0c2020, 0x0c0bdfe0, 0x0c0c2e2e, 0x0c0bd1d2, 0xf3f40000, 0xf3f40202, 0xf3f3fdfe, + 0xf3f40606, 0xf3f3f9fa, 0xf3f40c0c, 0xf3f3f3f4, 0xf3f41414, 0xf3f3ebec, 0xf3f42020, 0xf3f3dfe0, + 0xf3f42e2e, 0xf3f3d1d2, 0x14140000, 0x14140202, 0x1413fdfe, 0x14140606, 0x1413f9fa, 0x14140c0c, + 0x1413f3f4, 0x14141414, 0x1413ebec, 0x14142020, 0x1413dfe0, 0x14142e2e, 0x1413d1d2, 0xebec0000, + 0xebec0202, 0xebebfdfe, 0xebec0606, 0xebebf9fa, 0xebec0c0c, 0xebebf3f4, 0xebec1414, 0xebebebec, + 0xebec2020, 0xebebdfe0, 0xebec2e2e, 0xebebd1d2, 0x20200000, 0x20200202, 0x201ffdfe, 0x20200606, + 0x201ff9fa, 0x20200c0c, 0x201ff3f4, 0x20201414, 0x201febec, 0x20202020, 0x201fdfe0, 0x20202e2e, + 0x201fd1d2, 0xdfe00000, 0xdfe00202, 0xdfdffdfe, 0xdfe00606, 0xdfdff9fa, 0xdfe00c0c, 0xdfdff3f4, + 0xdfe01414, 0xdfdfebec, 0xdfe02020, 0xdfdfdfe0, 0xdfe02e2e, 0xdfdfd1d2, 0x2e2e0000, 0x2e2e0202, + 0x2e2dfdfe, 0x2e2e0606, 0x2e2df9fa, 0x2e2e0c0c, 0x2e2df3f4, 0x2e2e1414, 0x2e2debec, 0x2e2e2020, + 0x2e2ddfe0, 0x2e2e2e2e, 0x2e2dd1d2, 0xd1d20000, 0xd1d20202, 0xd1d1fdfe, 0xd1d20606, 0xd1d1f9fa, + 0xd1d20c0c, 0xd1d1f3f4, 0xd1d21414, 0xd1d1ebec, 0xd1d22020, 0xd1d1dfe0, 0xd1d22e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; static const uint32_t correctionloworder[] = { - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x04040404, - 0xfbfbfbfc, 0x05050101, 0xfafafeff, 0x01010505, 0xfefefafb, 0x0403fbfc, 0xfbfc0404, 0x0605fdfe, - 0xf9fa0202, 0xfdfe0606, 0x0201f9fa, 0x09090404, 0xf6f6fbfc, 0x04040909, 0xfbfbf6f7, 0x09090909, - 0xf6f6f6f7, 0x0a0a0101, 0xf5f5feff, 0x01010a0a, 0xfefef5f6, 0x0807fafb, 0xf7f80505, 0xfafb0808, - 0x0504f7f8, 0x0f0f0909, 0xf0f0f6f7, 0x09090f0f, 0xf6f6f0f1, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, - 0x0302f3f4, 0x10100404, 0xefeffbfc, 0x04041010, 0xfbfbeff0, 0x10101010, 0xefefeff0, 0x12120000, - 0xedee0000, 0x00001212, 0xffffedee, 0x0c0bf3f4, 0xf3f40c0c, 0x100ff6f7, 0xeff00909, 0xf6f71010, - 0x0908eff0, 0x1b1b0b0b, 0xe4e4f4f5, 0x0b0b1b1b, 0xf4f4e4e5, 0x1c1c1313, 0xe3e3eced, 0x13131c1c, - 0xecece3e4, 0x1615f9fa, 0xe9ea0606, 0xf9fa1616, 0x0605e9ea, 0x1d1d0404, 0xe2e2fbfc, 0x04041d1d, - 0xfbfbe2e3, 0x1e1e1e1e, 0xe1e1e1e2, 0x2120fdfe, 0xdedf0202, 0xfdfe2121, 0x0201dedf, 0x1716edee, - 0xe8e91212, 0xedee1717, 0x1211e8e9, 0x1e1df0f1, 0xe1e20f0f, 0xf0f11e1e, 0x0f0ee1e2, 0x2e2e1616, - 0xd1d1e9ea, 0x16162e2e, 0xe9e9d1d2, 0x2f2f0d0d, 0xd0d0f2f3, 0x0d0d2f2f, 0xf2f2d0d1, 0x31312323, - 0xcecedcdd, 0x23233131, 0xdcdccecf, 0x2928f4f5, 0xd6d70b0b, 0xf4f52929, 0x0b0ad6d7, 0x33330404, - 0xccccfbfc, 0x04043333, 0xfbfbcccd, 0x36363636, 0xc9c9c9ca, 0x2221ddde, 0xddde2222, 0x2a29e2e3, - 0xd5d61d1d, 0xe2e32a2a, 0x1d1cd5d6, 0x3c3bf9fa, 0xc3c40606, 0xf9fa3c3c, 0x0605c3c4, 0x4c4c1b1b, - 0xb3b3e4e5, 0x1b1b4c4c, 0xe4e4b3b4, 0x4d4d2b2b, 0xb2b2d4d5, 0x2b2b4d4d, 0xd4d4b2b3, 0x3736e7e8, - 0xc8c91818, 0xe7e83737, 0x1817c8c9, 0x4f4f0e0e, 0xb0b0f1f2, 0x0e0e4f4f, 0xf1f1b0b1, 0x53533f3f, - 0xacacc0c1, 0x3f3f5353, 0xc0c0acad, 0x4a49ebec, 0xb5b61414, 0xebec4a4a, 0x1413b5b6, 0x58580202, - 0xa7a7fdfe, 0x02025858, 0xfdfda7a8, 0x5d5d5d5d, 0xa2a2a2a3, 0x3d3ccbcc, 0xc2c33434, 0xcbcc3d3d, - 0x3433c2c3, 0x78783434, 0x8787cbcc, 0x34347878, 0xcbcb8788, 0x4b4ad2d3, 0xb4b52d2d, 0xd2d34b4b, - 0x2d2cb4b5, 0x7d7d4b4b, 0x8282b4b5, 0x4b4b7d7d, 0xb4b48283, 0x7a7a2121, 0x8585dedf, 0x21217a7a, - 0xdede8586, 0x6766f2f3, 0x98990d0d, 0xf2f36767, 0x0d0c9899, 0x605fd7d8, 0x9fa02828, 0xd7d86060, - 0x28279fa0, 0x7f7eddde, 0x80812222, 0xddde7f7f, 0x22218081, 0x5958a6a7, 0xa6a75959, 0x6968b1b2, - 0x96974e4e, 0xb1b26969, 0x4e4d9697, 0x0c0c0c0c, 0xf3f3f3f4, 0x17171717, 0xe8e8e8e9, 0x2a2a2a2a, - 0xd5d5d5d6, 0x49494949, 0xb6b6b6b7, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0xfcfd0101, - 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfeff0303, 0xfeff0303, - 0xfeff0303, 0xfeff0303, 0xfeff0303, 0xfeff0303, 0xfeff0303, 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, - 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, - 0xf8f8f8f9, 0x08080202, 0xf7f7fdfe, 0x02020808, 0xfdfdf7f8, 0x0908fdfe, 0xf6f70202, 0xfdfe0909, - 0x0201f6f7, 0x0605f9fa, 0xf9fa0606, 0x0d0d0606, 0xf2f2f9fa, 0x06060d0d, 0xf9f9f2f3, 0x0d0d0d0d, - 0xf2f2f2f3, 0x0e0e0101, 0xf1f1feff, 0x01010e0e, 0xfefef1f2, 0x0c0bf7f8, 0xf3f40808, 0xf7f80c0c, - 0x0807f3f4, 0x17170e0e, 0xe8e8f1f2, 0x0e0e1717, 0xf1f1e8e9, 0x1211fafb, 0xedee0505, 0xfafb1212, - 0x0504edee, 0x18180606, 0xe7e7f9fa, 0x06061818, 0xf9f9e7e8, 0x18181818, 0xe7e7e7e8, 0x1b1afeff, - 0xe4e50101, 0xfeff1b1b, 0x0100e4e5, 0x1110eeef, 0xeeef1111, 0x1716f2f3, 0xe8e90d0d, 0xf2f31717, - 0x0d0ce8e9, 0x28281010, 0xd7d7eff0, 0x10102828, 0xefefd7d8, 0x29291c1c, 0xd6d6e3e4, 0x1c1c2929, - 0xe3e3d6d7, 0x2120f6f7, 0xdedf0909, 0xf6f72121, 0x0908dedf, 0x2b2b0606, 0xd4d4f9fa, 0x06062b2b, - 0xf9f9d4d5, 0x2e2e2e2e, 0xd1d1d1d2, 0x3231fbfc, 0xcdce0404, 0xfbfc3232, 0x0403cdce, 0x2221e4e5, - 0xddde1b1b, 0xe4e52222, 0x1b1addde, 0x2d2ce9ea, 0xd2d31616, 0xe9ea2d2d, 0x1615d2d3, 0x45452222, - 0xbabaddde, 0x22224545, 0xddddbabb, 0x46461313, 0xb9b9eced, 0x13134646, 0xececb9ba, 0x49493535, - 0xb6b6cacb, 0x35354949, 0xcacab6b7, 0x3e3deeef, 0xc1c21111, 0xeeef3e3e, 0x1110c1c2, 0x4d4d0505, - 0xb2b2fafb, 0x05054d4d, 0xfafab2b3, 0x52525252, 0xadadadae, 0x3332cccd, 0xcccd3333, 0x403fd4d5, - 0xbfc02b2b, 0xd4d54040, 0x2b2abfc0, 0x5a59f5f6, 0xa5a60a0a, 0xf5f65a5a, 0x0a09a5a6, 0x72722929, - 0x8d8dd6d7, 0x29297272, 0xd6d68d8e, 0x74744040, 0x8b8bbfc0, 0x40407474, 0xbfbf8b8c, 0x5251dadb, - 0xadae2525, 0xdadb5252, 0x2524adae, 0x77771616, 0x8888e9ea, 0x16167777, 0xe9e98889, 0x7c7c5f5f, - 0x8383a0a1, 0x5f5f7c7c, 0xa0a08384, 0x6f6ee1e2, 0x90911e1e, 0xe1e26f6f, 0x1e1d9091, 0x5c5bb1b2, - 0xa3a44e4e, 0xb1b25c5c, 0x4e4da3a4, 0x7170bbbc, 0x8e8f4444, 0xbbbc7171, 0x44438e8f, 0x12121212, - 0xedededee, 0x22222222, 0xddddddde, 0x3f3f3f3f, 0xc0c0c0c1, 0x6d6d6d6d, 0x92929293, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, - 0x03030303, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, - 0xfcfcfcfd, 0xfcfcfcfd, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, - 0x0403feff, 0x0403feff, 0x0403feff, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, - 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, - 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, - 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x07070707, 0x07070707, - 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, - 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, - 0xf5f5fcfd, 0x03030a0a, 0xfcfcf5f6, 0x09090909, 0xf6f6f6f7, 0x0706f8f9, 0xf8f90707, 0x0c0bfcfd, - 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x11110808, 0xeeeef7f8, 0x08081111, 0xf7f7eeef, 0x11111111, - 0xeeeeeeef, 0x13130101, 0xececfeff, 0x01011313, 0xfefeeced, 0x100ff4f5, 0xeff00b0b, 0xf4f51010, - 0x0b0aeff0, 0x1716f9fa, 0xe8e90606, 0xf9fa1717, 0x0605e8e9, 0x1f1f1212, 0xe0e0edee, 0x12121f1f, - 0xedede0e1, 0x20200808, 0xdfdff7f8, 0x08082020, 0xf7f7dfe0, 0x21212121, 0xdedededf, 0x2423feff, - 0xdbdc0101, 0xfeff2424, 0x0100dbdc, 0x1716e8e9, 0xe8e91717, 0x1f1eeeef, 0xe0e11111, 0xeeef1f1f, - 0x1110e0e1, 0x36361515, 0xc9c9eaeb, 0x15153636, 0xeaeac9ca, 0x37372525, 0xc8c8dadb, 0x25253737, - 0xdadac8c9, 0x2c2bf3f4, 0xd3d40c0c, 0xf3f42c2c, 0x0c0bd3d4, 0x39390808, 0xc6c6f7f8, 0x08083939, - 0xf7f7c6c7, 0x3d3d3d3d, 0xc2c2c2c3, 0x4241fafb, 0xbdbe0505, 0xfafb4242, 0x0504bdbe, 0x2d2cdbdc, - 0xd2d32424, 0xdbdc2d2d, 0x2423d2d3, 0x3c3be2e3, 0xc3c41d1d, 0xe2e33c3c, 0x1d1cc3c4, 0x5c5c2d2d, - 0xa3a3d2d3, 0x2d2d5c5c, 0xd2d2a3a4, 0x5d5d1919, 0xa2a2e6e7, 0x19195d5d, 0xe6e6a2a3, 0x61614747, - 0x9e9eb8b9, 0x47476161, 0xb8b89e9f, 0x5352e9ea, 0xacad1616, 0xe9ea5353, 0x1615acad, 0x66660707, - 0x9999f8f9, 0x07076666, 0xf8f8999a, 0x6d6d6d6d, 0x92929293, 0x4443bbbc, 0xbbbc4444, 0x5554c6c7, - 0xaaab3939, 0xc6c75555, 0x3938aaab, 0x7877f2f3, 0x87880d0d, 0xf2f37878, 0x0d0c8788, 0x6e6dcecf, - 0x91923131, 0xcecf6e6e, 0x31309192, 0x7b7a9798, 0x84856868, 0x97987b7b, 0x68678485, 0x18181818, - 0xe7e7e7e8, 0x2e2e2e2e, 0xd1d1d1d2, 0x54545454, 0xabababac, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040404, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, - 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, - 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0xfafb0101, 0xfafb0101, 0xfafb0101, - 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfeff0505, - 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, - 0xfeff0505, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, - 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, - 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, - 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0x03030a0a, - 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, - 0x03030a0a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, - 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x03030d0d, 0xfcfcf2f3, 0x0908f6f7, 0xf6f70909, 0x0f0efbfc, - 0xf0f10404, 0xfbfc0f0f, 0x0403f0f1, 0x16160b0b, 0xe9e9f4f5, 0x0b0b1616, 0xf4f4e9ea, 0x15151515, - 0xeaeaeaeb, 0x18180202, 0xe7e7fdfe, 0x02021818, 0xfdfde7e8, 0x1413f1f2, 0xebec0e0e, 0xf1f21414, - 0x0e0debec, 0x26261717, 0xd9d9e8e9, 0x17172626, 0xe8e8d9da, 0x1d1cf7f8, 0xe2e30808, 0xf7f81d1d, - 0x0807e2e3, 0x27270b0b, 0xd8d8f4f5, 0x0b0b2727, 0xf4f4d8d9, 0x29292929, 0xd6d6d6d7, 0x2d2cfeff, - 0xd2d30101, 0xfeff2d2d, 0x0100d2d3, 0x1d1ce2e3, 0xe2e31d1d, 0x2726e9ea, 0xd8d91616, 0xe9ea2727, - 0x1615d8d9, 0x43431b1b, 0xbcbce4e5, 0x1b1b4343, 0xe4e4bcbd, 0x45452f2f, 0xbabad0d1, 0x2f2f4545, - 0xd0d0babb, 0x3837f0f1, 0xc7c80f0f, 0xf0f13838, 0x0f0ec7c8, 0x47470b0b, 0xb8b8f4f5, 0x0b0b4747, - 0xf4f4b8b9, 0x4c4c4c4c, 0xb3b3b3b4, 0x5352f9fa, 0xacad0606, 0xf9fa5353, 0x0605acad, 0x3938d2d3, - 0xc6c72d2d, 0xd2d33939, 0x2d2cc6c7, 0x4b4adbdc, 0xb4b52424, 0xdbdc4b4b, 0x2423b4b5, 0x73733838, - 0x8c8cc7c8, 0x38387373, 0xc7c78c8d, 0x75751f1f, 0x8a8ae0e1, 0x1f1f7575, 0xe0e08a8b, 0x7a7a5858, - 0x8585a7a8, 0x58587a7a, 0xa7a78586, 0x6867e3e4, 0x97981c1c, 0xe3e46868, 0x1c1b9798, 0x5554aaab, - 0xaaab5555, 0x6a69b7b8, 0x95964848, 0xb7b86a6a, 0x48479596, 0x1e1e1e1e, 0xe1e1e1e2, 0x3a3a3a3a, - 0xc5c5c5c6, 0x69696969, 0x96969697, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x05050505, 0x05050505, - 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, - 0x05050505, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, - 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, - 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0xf8f90202, - 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, - 0xf8f90202, 0xf8f90202, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, - 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, - 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, - 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, - 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, - 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0x0d0d0303, 0x0d0d0303, - 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, - 0x0d0d0303, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, - 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, - 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0xfbfbf0f1, 0x0b0af4f5, 0xf4f50b0b, 0x1211fafb, - 0xedee0505, 0xfafb1212, 0x0504edee, 0x1a1a0d0d, 0xe5e5f2f3, 0x0d0d1a1a, 0xf2f2e5e6, 0x1a1a1a1a, - 0xe5e5e5e6, 0x1d1d0202, 0xe2e2fdfe, 0x02021d1d, 0xfdfde2e3, 0x1817eff0, 0xe7e81010, 0xeff01818, - 0x100fe7e8, 0x2e2e1c1c, 0xd1d1e3e4, 0x1c1c2e2e, 0xe3e3d1d2, 0x2322f6f7, 0xdcdd0909, 0xf6f72323, - 0x0908dcdd, 0x2f2f0d0d, 0xd0d0f2f3, 0x0d0d2f2f, 0xf2f2d0d1, 0x31313131, 0xcecececf, 0x3635feff, - 0xc9ca0101, 0xfeff3636, 0x0100c9ca, 0x2322dcdd, 0xdcdd2323, 0x2f2ee5e6, 0xd0d11a1a, 0xe5e62f2f, - 0x1a19d0d1, 0x51512020, 0xaeaedfe0, 0x20205151, 0xdfdfaeaf, 0x53533838, 0xacacc7c8, 0x38385353, - 0xc7c7acad, 0x4342edee, 0xbcbd1212, 0xedee4343, 0x1211bcbd, 0x56560d0d, 0xa9a9f2f3, 0x0d0d5656, - 0xf2f2a9aa, 0x5b5b5b5b, 0xa4a4a4a5, 0x6362f8f9, 0x9c9d0707, 0xf8f96363, 0x07069c9d, 0x4443c9ca, - 0xbbbc3636, 0xc9ca4444, 0x3635bbbc, 0x5a59d3d4, 0xa5a62c2c, 0xd3d45a5a, 0x2c2ba5a6, 0x7c7bdedf, - 0x83842121, 0xdedf7c7c, 0x21208384, 0x67669899, 0x98996767, 0x7f7ea9aa, 0x80815656, 0xa9aa7f7f, - 0x56558081, 0x25252525, 0xdadadadb, 0x45454545, 0xbabababb, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, - 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0xf7f80202, 0xf7f80202, 0xf7f80202, - 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, - 0xf7f80202, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, - 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, - 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, - 0x0201f7f8, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, - 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, - 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, - 0xf2f2f2f3, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, - 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, - 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, - 0xf0f0fbfc, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, - 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, - 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0xfafaedee, 0x0d0cf2f3, 0xf2f30d0d, 0x1514f9fa, - 0xeaeb0606, 0xf9fa1515, 0x0605eaeb, 0x1e1e0f0f, 0xe1e1f0f1, 0x0f0f1e1e, 0xf0f0e1e2, 0x1e1e1e1e, - 0xe1e1e1e2, 0x22220202, 0xddddfdfe, 0x02022222, 0xfdfdddde, 0x1c1beced, 0xe3e41313, 0xeced1c1c, - 0x1312e3e4, 0x36362020, 0xc9c9dfe0, 0x20203636, 0xdfdfc9ca, 0x2928f4f5, 0xd6d70b0b, 0xf4f52929, - 0x0b0ad6d7, 0x37370f0f, 0xc8c8f0f1, 0x0f0f3737, 0xf0f0c8c9, 0x39393939, 0xc6c6c6c7, 0x3f3efeff, - 0xc0c10101, 0xfeff3f3f, 0x0100c0c1, 0x2827d7d8, 0xd7d82828, 0x3736e1e2, 0xc8c91e1e, 0xe1e23737, - 0x1e1dc8c9, 0x5e5e2525, 0xa1a1dadb, 0x25255e5e, 0xdadaa1a2, 0x60604141, 0x9f9fbebf, 0x41416060, - 0xbebe9fa0, 0x4e4deaeb, 0xb1b21515, 0xeaeb4e4e, 0x1514b1b2, 0x64640f0f, 0x9b9bf0f1, 0x0f0f6464, - 0xf0f09b9c, 0x6a6a6a6a, 0x95959596, 0x7473f7f8, 0x8b8c0808, 0xf7f87474, 0x08078b8c, 0x4f4ec0c1, - 0xb0b13f3f, 0xc0c14f4f, 0x3f3eb0b1, 0x6968cccd, 0x96973333, 0xcccd6969, 0x33329697, 0x78778788, - 0x87887878, 0x2b2b2b2b, 0xd4d4d4d5, 0x50505050, 0xafafafb0, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, - 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, - 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, - 0xf8f8f8f9, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, - 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0xf5f60303, 0xf5f60303, 0xf5f60303, - 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, - 0xf5f60303, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, - 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, - 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, - 0x0302f5f6, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, - 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0xefefeff0, 0xefefeff0, 0xefefeff0, - 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, - 0xefefeff0, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, - 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0xededfafb, 0xededfafb, 0xededfafb, - 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, - 0xededfafb, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, - 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, - 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0xfafaebec, 0x0f0ef0f1, 0xf0f10f0f, 0x1817f8f9, - 0xe7e80707, 0xf8f91818, 0x0706e7e8, 0x23231111, 0xdcdceeef, 0x11112323, 0xeeeedcdd, 0x22222222, - 0xddddddde, 0x26260303, 0xd9d9fcfd, 0x03032626, 0xfcfcd9da, 0x201fe9ea, 0xdfe01616, 0xe9ea2020, - 0x1615dfe0, 0x3d3d2525, 0xc2c2dadb, 0x25253d3d, 0xdadac2c3, 0x2f2ef2f3, 0xd0d10d0d, 0xf2f32f2f, - 0x0d0cd0d1, 0x3f3f1111, 0xc0c0eeef, 0x11113f3f, 0xeeeec0c1, 0x41414141, 0xbebebebf, 0x4847feff, - 0xb7b80101, 0xfeff4848, 0x0100b7b8, 0x2e2dd1d2, 0xd1d22e2e, 0x3f3edcdd, 0xc0c12323, 0xdcdd3f3f, - 0x2322c0c1, 0x6b6b2b2b, 0x9494d4d5, 0x2b2b6b6b, 0xd4d49495, 0x6e6e4b4b, 0x9191b4b5, 0x4b4b6e6e, - 0xb4b49192, 0x5958e7e8, 0xa6a71818, 0xe7e85959, 0x1817a6a7, 0x72721111, 0x8d8deeef, 0x11117272, - 0xeeee8d8e, 0x79797979, 0x86868687, 0x5b5ab7b8, 0xa4a54848, 0xb7b85b5b, 0x4847a4a5, 0x7877c5c6, - 0x87883a3a, 0xc5c67878, 0x3a398788, 0x31313131, 0xcecececf, 0x5c5c5c5c, 0xa3a3a3a4, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, - 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0xf7f7f7f8, - 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, - 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, - 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0xf4f50303, - 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, - 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, - 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0x0302f4f5, - 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, - 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, - 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0xedededee, - 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, - 0xedededee, 0xedededee, 0xedededee, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, - 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0xebebfafb, - 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, - 0xebebfafb, 0xebebfafb, 0xebebfafb, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, - 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, - 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x1110eeef, 0xeeef1111, 0x1b1af8f9, - 0xe4e50707, 0xf8f91b1b, 0x0706e4e5, 0x27271313, 0xd8d8eced, 0x13132727, 0xececd8d9, 0x27272727, - 0xd8d8d8d9, 0x2b2b0303, 0xd4d4fcfd, 0x03032b2b, 0xfcfcd4d5, 0x2423e7e8, 0xdbdc1818, 0xe7e82424, - 0x1817dbdc, 0x45452a2a, 0xbabad5d6, 0x2a2a4545, 0xd5d5babb, 0x3534f1f2, 0xcacb0e0e, 0xf1f23535, - 0x0e0dcacb, 0x47471313, 0xb8b8eced, 0x13134747, 0xececb8b9, 0x49494949, 0xb6b6b6b7, 0x504ffdfe, - 0xafb00202, 0xfdfe5050, 0x0201afb0, 0x3433cbcc, 0xcbcc3434, 0x4645d8d9, 0xb9ba2727, 0xd8d94646, - 0x2726b9ba, 0x79793030, 0x8686cfd0, 0x30307979, 0xcfcf8687, 0x7c7c5454, 0x8383abac, 0x54547c7c, - 0xabab8384, 0x6463e4e5, 0x9b9c1b1b, 0xe4e56464, 0x1b1a9b9c, 0x6665aeaf, 0x999a5151, 0xaeaf6666, - 0x5150999a, 0x37373737, 0xc8c8c8c9, 0x68686868, 0x97979798, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, - 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0xf6f6f6f7, - 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, - 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, - 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, - 0x0c0bfcfd, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, - 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xfcfd0c0c, 0xfcfd0c0c, - 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, - 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, - 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, - 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0xe8e8f9fa, - 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, - 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0x06061717, 0x06061717, 0x06061717, 0x06061717, - 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, - 0x06061717, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, - 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, - 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x0403fbfc, 0xfbfc0404, 0x0605fdfe, - 0xf9fa0202, 0xfdfe0606, 0x0201f9fa, 0x08080404, 0xf7f7fbfc, 0x04040808, 0xfbfbf7f8, 0x08080808, - 0xf7f7f7f8, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x0807fbfc, 0xf7f80404, 0xfbfc0808, - 0x0403f7f8, 0x0e0e0808, 0xf1f1f7f8, 0x08080e0e, 0xf7f7f1f2, 0x0c0bfdfe, 0xf3f40202, 0xfdfe0c0c, - 0x0201f3f4, 0x10100404, 0xefeffbfc, 0x04041010, 0xfbfbeff0, 0x10101010, 0xefefeff0, 0x12120000, - 0xedee0000, 0x00001212, 0xffffedee, 0x0c0bf3f4, 0xf3f40c0c, 0x100ff7f8, 0xeff00808, 0xf7f81010, - 0x0807eff0, 0x1a1a0a0a, 0xe5e5f5f6, 0x0a0a1a1a, 0xf5f5e5e6, 0x1c1c1212, 0xe3e3edee, 0x12121c1c, - 0xedede3e4, 0x1615f9fa, 0xe9ea0606, 0xf9fa1616, 0x0605e9ea, 0x1c1c0404, 0xe3e3fbfc, 0x04041c1c, - 0xfbfbe3e4, 0x1e1e1e1e, 0xe1e1e1e2, 0x201ffdfe, 0xdfe00202, 0xfdfe2020, 0x0201dfe0, 0x1615edee, - 0xe9ea1212, 0xedee1616, 0x1211e9ea, 0x1e1df1f2, 0xe1e20e0e, 0xf1f21e1e, 0x0e0de1e2, 0x2e2e1616, - 0xd1d1e9ea, 0x16162e2e, 0xe9e9d1d2, 0x2e2e0c0c, 0xd1d1f3f4, 0x0c0c2e2e, 0xf3f3d1d2, 0x30302222, - 0xcfcfddde, 0x22223030, 0xddddcfd0, 0x2827f5f6, 0xd7d80a0a, 0xf5f62828, 0x0a09d7d8, 0x32320404, - 0xcdcdfbfc, 0x04043232, 0xfbfbcdce, 0x36363636, 0xc9c9c9ca, 0x2221ddde, 0xddde2222, 0x2a29e3e4, - 0xd5d61c1c, 0xe3e42a2a, 0x1c1bd5d6, 0x3c3bf9fa, 0xc3c40606, 0xf9fa3c3c, 0x0605c3c4, 0x4c4c1a1a, - 0xb3b3e5e6, 0x1a1a4c4c, 0xe5e5b3b4, 0x4c4c2a2a, 0xb3b3d5d6, 0x2a2a4c4c, 0xd5d5b3b4, 0x3635e7e8, - 0xc9ca1818, 0xe7e83636, 0x1817c9ca, 0x4e4e0e0e, 0xb1b1f1f2, 0x0e0e4e4e, 0xf1f1b1b2, 0x52523e3e, - 0xadadc1c2, 0x3e3e5252, 0xc1c1adae, 0x4a49ebec, 0xb5b61414, 0xebec4a4a, 0x1413b5b6, 0x58580202, - 0xa7a7fdfe, 0x02025858, 0xfdfda7a8, 0x5c5c5c5c, 0xa3a3a3a4, 0x3c3bcbcc, 0xc3c43434, 0xcbcc3c3c, - 0x3433c3c4, 0x76763434, 0x8989cbcc, 0x34347676, 0xcbcb898a, 0x4a49d3d4, 0xb5b62c2c, 0xd3d44a4a, - 0x2c2bb5b6, 0x76764a4a, 0x8989b5b6, 0x4a4a7676, 0xb5b5898a, 0x76762020, 0x8989dfe0, 0x20207676, - 0xdfdf898a, 0x6665f3f4, 0x999a0c0c, 0xf3f46666, 0x0c0b999a, 0x605fd7d8, 0x9fa02828, 0xd7d86060, - 0x28279fa0, 0x7675ddde, 0x898a2222, 0xddde7676, 0x2221898a, 0x5857a7a8, 0xa7a85858, 0x6867b1b2, - 0x97984e4e, 0xb1b26868, 0x4e4d9798, 0x0c0c0c0c, 0xf3f3f3f4, 0x16161616, 0xe9e9e9ea, 0x2a2a2a2a, - 0xd5d5d5d6, 0x48484848, 0xb7b7b7b8, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0xfdfe0000, - 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, - 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, - 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, - 0xf9f9f9fa, 0x09090303, 0xf6f6fcfd, 0x03030909, 0xfcfcf6f7, 0x0908fcfd, 0xf6f70303, 0xfcfd0909, - 0x0302f6f7, 0x0605f9fa, 0xf9fa0606, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0xf9f9f3f4, 0x0c0c0c0c, - 0xf3f3f3f4, 0x0f0f0000, 0xf0f10000, 0x00000f0f, 0xfffff0f1, 0x0c0bf6f7, 0xf3f40909, 0xf6f70c0c, - 0x0908f3f4, 0x18180f0f, 0xe7e7f0f1, 0x0f0f1818, 0xf0f0e7e8, 0x1211f9fa, 0xedee0606, 0xf9fa1212, - 0x0605edee, 0x18180606, 0xe7e7f9fa, 0x06061818, 0xf9f9e7e8, 0x18181818, 0xe7e7e7e8, 0x1b1b0000, - 0xe4e50000, 0x00001b1b, 0xffffe4e5, 0x1211edee, 0xedee1212, 0x1817f3f4, 0xe7e80c0c, 0xf3f41818, - 0x0c0be7e8, 0x27270f0f, 0xd8d8f0f1, 0x0f0f2727, 0xf0f0d8d9, 0x2a2a1b1b, 0xd5d5e4e5, 0x1b1b2a2a, - 0xe4e4d5d6, 0x2120f6f7, 0xdedf0909, 0xf6f72121, 0x0908dedf, 0x2a2a0606, 0xd5d5f9fa, 0x06062a2a, - 0xf9f9d5d6, 0x2d2d2d2d, 0xd2d2d2d3, 0x3332fcfd, 0xcccd0303, 0xfcfd3333, 0x0302cccd, 0x2120e4e5, - 0xdedf1b1b, 0xe4e52121, 0x1b1adedf, 0x2d2ceaeb, 0xd2d31515, 0xeaeb2d2d, 0x1514d2d3, 0x45452121, - 0xbabadedf, 0x21214545, 0xdedebabb, 0x45451212, 0xbabaedee, 0x12124545, 0xededbabb, 0x48483636, - 0xb7b7c9ca, 0x36364848, 0xc9c9b7b8, 0x3f3eedee, 0xc0c11212, 0xedee3f3f, 0x1211c0c1, 0x4e4e0606, - 0xb1b1f9fa, 0x06064e4e, 0xf9f9b1b2, 0x51515151, 0xaeaeaeaf, 0x3332cccd, 0xcccd3333, 0x3f3ed5d6, - 0xc0c12a2a, 0xd5d63f3f, 0x2a29c0c1, 0x5a59f6f7, 0xa5a60909, 0xf6f75a5a, 0x0908a5a6, 0x72722a2a, - 0x8d8dd5d6, 0x2a2a7272, 0xd5d58d8e, 0x75753f3f, 0x8a8ac0c1, 0x3f3f7575, 0xc0c08a8b, 0x5150dbdc, - 0xaeaf2424, 0xdbdc5151, 0x2423aeaf, 0x78781515, 0x8787eaeb, 0x15157878, 0xeaea8788, 0x7b7b6060, - 0x84849fa0, 0x60607b7b, 0x9f9f8485, 0x6f6ee1e2, 0x90911e1e, 0xe1e26f6f, 0x1e1d9091, 0x5d5cb1b2, - 0xa2a34e4e, 0xb1b25d5d, 0x4e4da2a3, 0x7271babb, 0x8d8e4545, 0xbabb7272, 0x45448d8e, 0x12121212, - 0xedededee, 0x21212121, 0xdedededf, 0x3f3f3f3f, 0xc0c0c0c1, 0x6c6c6c6c, 0x93939394, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, - 0x03030303, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, - 0xfcfcfcfd, 0xfcfcfcfd, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, - 0x03030000, 0x03030000, 0x03030000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, - 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0x00000303, 0x00000303, 0x00000303, 0x00000303, - 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, - 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, - 0xf7f7fbfc, 0x04040808, 0xfbfbf7f8, 0x08080808, 0xf7f7f7f8, 0x0807f7f8, 0xf7f80808, 0x0c0bfbfc, - 0xf3f40404, 0xfbfc0c0c, 0x0403f3f4, 0x10100808, 0xefeff7f8, 0x08081010, 0xf7f7eff0, 0x10101010, - 0xefefeff0, 0x14140000, 0xebec0000, 0x00001414, 0xffffebec, 0x100ff3f4, 0xeff00c0c, 0xf3f41010, - 0x0c0beff0, 0x1817fbfc, 0xe7e80404, 0xfbfc1818, 0x0403e7e8, 0x20201010, 0xdfdfeff0, 0x10102020, - 0xefefdfe0, 0x20200808, 0xdfdff7f8, 0x08082020, 0xf7f7dfe0, 0x20202020, 0xdfdfdfe0, 0x24240000, - 0xdbdc0000, 0x00002424, 0xffffdbdc, 0x1817e7e8, 0xe7e81818, 0x201feff0, 0xdfe01010, 0xeff02020, - 0x100fdfe0, 0x34341414, 0xcbcbebec, 0x14143434, 0xebebcbcc, 0x38382424, 0xc7c7dbdc, 0x24243838, - 0xdbdbc7c8, 0x2c2bf3f4, 0xd3d40c0c, 0xf3f42c2c, 0x0c0bd3d4, 0x38380808, 0xc7c7f7f8, 0x08083838, - 0xf7f7c7c8, 0x3c3c3c3c, 0xc3c3c3c4, 0x403ffbfc, 0xbfc00404, 0xfbfc4040, 0x0403bfc0, 0x2c2bdbdc, - 0xd3d42424, 0xdbdc2c2c, 0x2423d3d4, 0x3c3be3e4, 0xc3c41c1c, 0xe3e43c3c, 0x1c1bc3c4, 0x5c5c2c2c, - 0xa3a3d3d4, 0x2c2c5c5c, 0xd3d3a3a4, 0x5c5c1818, 0xa3a3e7e8, 0x18185c5c, 0xe7e7a3a4, 0x60604848, - 0x9f9fb7b8, 0x48486060, 0xb7b79fa0, 0x5453ebec, 0xabac1414, 0xebec5454, 0x1413abac, 0x64640808, - 0x9b9bf7f8, 0x08086464, 0xf7f79b9c, 0x6c6c6c6c, 0x93939394, 0x4443bbbc, 0xbbbc4444, 0x5453c7c8, - 0xabac3838, 0xc7c85454, 0x3837abac, 0x7877f3f4, 0x87880c0c, 0xf3f47878, 0x0c0b8788, 0x6c6bcfd0, - 0x93943030, 0xcfd06c6c, 0x302f9394, 0x7c7b9798, 0x83846868, 0x97987c7c, 0x68678384, 0x18181818, - 0xe7e7e7e8, 0x2c2c2c2c, 0xd3d3d3d4, 0x54545454, 0xabababac, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040404, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, - 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, - 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, - 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0x00000404, - 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, - 0x00000404, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, - 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0x08080404, - 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, - 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0x04040808, - 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, - 0x04040808, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, - 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x05050f0f, 0xfafaf0f1, 0x0a09f5f6, 0xf5f60a0a, 0x0f0efafb, - 0xf0f10505, 0xfafb0f0f, 0x0504f0f1, 0x14140a0a, 0xebebf5f6, 0x0a0a1414, 0xf5f5ebec, 0x14141414, - 0xebebebec, 0x19190000, 0xe6e70000, 0x00001919, 0xffffe6e7, 0x1413f0f1, 0xebec0f0f, 0xf0f11414, - 0x0f0eebec, 0x28281919, 0xd7d7e6e7, 0x19192828, 0xe6e6d7d8, 0x1e1df5f6, 0xe1e20a0a, 0xf5f61e1e, - 0x0a09e1e2, 0x28280a0a, 0xd7d7f5f6, 0x0a0a2828, 0xf5f5d7d8, 0x28282828, 0xd7d7d7d8, 0x2d2d0000, - 0xd2d30000, 0x00002d2d, 0xffffd2d3, 0x1e1de1e2, 0xe1e21e1e, 0x2827ebec, 0xd7d81414, 0xebec2828, - 0x1413d7d8, 0x41411919, 0xbebee6e7, 0x19194141, 0xe6e6bebf, 0x46462d2d, 0xb9b9d2d3, 0x2d2d4646, - 0xd2d2b9ba, 0x3736f0f1, 0xc8c90f0f, 0xf0f13737, 0x0f0ec8c9, 0x46460a0a, 0xb9b9f5f6, 0x0a0a4646, - 0xf5f5b9ba, 0x4b4b4b4b, 0xb4b4b4b5, 0x5554fafb, 0xaaab0505, 0xfafb5555, 0x0504aaab, 0x3736d2d3, - 0xc8c92d2d, 0xd2d33737, 0x2d2cc8c9, 0x4b4adcdd, 0xb4b52323, 0xdcdd4b4b, 0x2322b4b5, 0x73733737, - 0x8c8cc8c9, 0x37377373, 0xc8c88c8d, 0x73731e1e, 0x8c8ce1e2, 0x1e1e7373, 0xe1e18c8d, 0x78785a5a, - 0x8787a5a6, 0x5a5a7878, 0xa5a58788, 0x6968e1e2, 0x96971e1e, 0xe1e26969, 0x1e1d9697, 0x5554aaab, - 0xaaab5555, 0x6968b9ba, 0x96974646, 0xb9ba6969, 0x46459697, 0x1e1e1e1e, 0xe1e1e1e2, 0x3c3c3c3c, - 0xc3c3c3c4, 0x69696969, 0x96969697, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x05050505, 0x05050505, - 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, - 0x05050505, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, - 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0x05050000, 0x05050000, 0x05050000, 0x05050000, - 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0xfafb0000, - 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, - 0xfafb0000, 0xfafb0000, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, - 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0xfffffafb, 0xfffffafb, 0xfffffafb, - 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, - 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, - 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, - 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0x0f0f0505, 0x0f0f0505, - 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, - 0x0f0f0505, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, - 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, - 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0xf9f9f3f4, 0x0c0bf3f4, 0xf3f40c0c, 0x1211f9fa, - 0xedee0606, 0xf9fa1212, 0x0605edee, 0x18180c0c, 0xe7e7f3f4, 0x0c0c1818, 0xf3f3e7e8, 0x18181818, - 0xe7e7e7e8, 0x1e1e0000, 0xe1e20000, 0x00001e1e, 0xffffe1e2, 0x1817edee, 0xe7e81212, 0xedee1818, - 0x1211e7e8, 0x30301e1e, 0xcfcfe1e2, 0x1e1e3030, 0xe1e1cfd0, 0x2423f9fa, 0xdbdc0606, 0xf9fa2424, - 0x0605dbdc, 0x30300c0c, 0xcfcff3f4, 0x0c0c3030, 0xf3f3cfd0, 0x30303030, 0xcfcfcfd0, 0x36360000, - 0xc9ca0000, 0x00003636, 0xffffc9ca, 0x2423dbdc, 0xdbdc2424, 0x302fe7e8, 0xcfd01818, 0xe7e83030, - 0x1817cfd0, 0x4e4e1e1e, 0xb1b1e1e2, 0x1e1e4e4e, 0xe1e1b1b2, 0x54543636, 0xababc9ca, 0x36365454, - 0xc9c9abac, 0x4241edee, 0xbdbe1212, 0xedee4242, 0x1211bdbe, 0x54540c0c, 0xababf3f4, 0x0c0c5454, - 0xf3f3abac, 0x5a5a5a5a, 0xa5a5a5a6, 0x605ff9fa, 0x9fa00606, 0xf9fa6060, 0x06059fa0, 0x4241c9ca, - 0xbdbe3636, 0xc9ca4242, 0x3635bdbe, 0x5a59d5d6, 0xa5a62a2a, 0xd5d65a5a, 0x2a29a5a6, 0x7e7de1e2, - 0x81821e1e, 0xe1e27e7e, 0x1e1d8182, 0x6665999a, 0x999a6666, 0x7e7dabac, 0x81825454, 0xabac7e7e, - 0x54538182, 0x24242424, 0xdbdbdbdc, 0x42424242, 0xbdbdbdbe, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, - 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, - 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, - 0xf9fa0000, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, - 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, - 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, - 0xfffff9fa, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, - 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, - 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, - 0xf3f3f9fa, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, - 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, - 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0xf8f8eaeb, 0x0e0df1f2, 0xf1f20e0e, 0x1514f8f9, - 0xeaeb0707, 0xf8f91515, 0x0706eaeb, 0x1c1c0e0e, 0xe3e3f1f2, 0x0e0e1c1c, 0xf1f1e3e4, 0x1c1c1c1c, - 0xe3e3e3e4, 0x23230000, 0xdcdd0000, 0x00002323, 0xffffdcdd, 0x1c1beaeb, 0xe3e41515, 0xeaeb1c1c, - 0x1514e3e4, 0x38382323, 0xc7c7dcdd, 0x23233838, 0xdcdcc7c8, 0x2a29f1f2, 0xd5d60e0e, 0xf1f22a2a, - 0x0e0dd5d6, 0x38380e0e, 0xc7c7f1f2, 0x0e0e3838, 0xf1f1c7c8, 0x38383838, 0xc7c7c7c8, 0x3f3f0000, - 0xc0c10000, 0x00003f3f, 0xffffc0c1, 0x2a29d5d6, 0xd5d62a2a, 0x3837e3e4, 0xc7c81c1c, 0xe3e43838, - 0x1c1bc7c8, 0x5b5b2323, 0xa4a4dcdd, 0x23235b5b, 0xdcdca4a5, 0x62623f3f, 0x9d9dc0c1, 0x3f3f6262, - 0xc0c09d9e, 0x4d4ceaeb, 0xb2b31515, 0xeaeb4d4d, 0x1514b2b3, 0x62620e0e, 0x9d9df1f2, 0x0e0e6262, - 0xf1f19d9e, 0x69696969, 0x96969697, 0x7776f8f9, 0x88890707, 0xf8f97777, 0x07068889, 0x4d4cc0c1, - 0xb2b33f3f, 0xc0c14d4d, 0x3f3eb2b3, 0x6968cecf, 0x96973131, 0xcecf6969, 0x31309697, 0x77768889, - 0x88897777, 0x2a2a2a2a, 0xd5d5d5d6, 0x4d4d4d4d, 0xb2b2b2b3, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, - 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, - 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, - 0xf8f8f8f9, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, - 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0xf8f90000, 0xf8f90000, 0xf8f90000, - 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, - 0xf8f90000, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, - 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, - 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, - 0xfffff8f9, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, - 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, - 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, - 0xf1f1f1f2, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, - 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, - 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, - 0xeaeaf8f9, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, - 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, - 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0xf7f7eff0, 0x100feff0, 0xeff01010, 0x1817f7f8, - 0xe7e80808, 0xf7f81818, 0x0807e7e8, 0x20201010, 0xdfdfeff0, 0x10102020, 0xefefdfe0, 0x20202020, - 0xdfdfdfe0, 0x28280000, 0xd7d80000, 0x00002828, 0xffffd7d8, 0x201fe7e8, 0xdfe01818, 0xe7e82020, - 0x1817dfe0, 0x40402828, 0xbfbfd7d8, 0x28284040, 0xd7d7bfc0, 0x302feff0, 0xcfd01010, 0xeff03030, - 0x100fcfd0, 0x40401010, 0xbfbfeff0, 0x10104040, 0xefefbfc0, 0x40404040, 0xbfbfbfc0, 0x48480000, - 0xb7b80000, 0x00004848, 0xffffb7b8, 0x302fcfd0, 0xcfd03030, 0x403fdfe0, 0xbfc02020, 0xdfe04040, - 0x201fbfc0, 0x68682828, 0x9797d7d8, 0x28286868, 0xd7d79798, 0x70704848, 0x8f8fb7b8, 0x48487070, - 0xb7b78f90, 0x5857e7e8, 0xa7a81818, 0xe7e85858, 0x1817a7a8, 0x70701010, 0x8f8feff0, 0x10107070, - 0xefef8f90, 0x78787878, 0x87878788, 0x5857b7b8, 0xa7a84848, 0xb7b85858, 0x4847a7a8, 0x7877c7c8, - 0x87883838, 0xc7c87878, 0x38378788, 0x30303030, 0xcfcfcfd0, 0x58585858, 0xa7a7a7a8, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, - 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0xf7f7f7f8, - 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, - 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, - 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0xf7f80000, - 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, - 0xf7f80000, 0xf7f80000, 0xf7f80000, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, - 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0xfffff7f8, - 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, - 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, - 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0xefefeff0, - 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, - 0xefefeff0, 0xefefeff0, 0xefefeff0, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, - 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0xefeff7f8, - 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, - 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, - 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, - 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x1211edee, 0xedee1212, 0x1b1af6f7, - 0xe4e50909, 0xf6f71b1b, 0x0908e4e5, 0x24241212, 0xdbdbedee, 0x12122424, 0xededdbdc, 0x24242424, - 0xdbdbdbdc, 0x2d2d0000, 0xd2d30000, 0x00002d2d, 0xffffd2d3, 0x2423e4e5, 0xdbdc1b1b, 0xe4e52424, - 0x1b1adbdc, 0x48482d2d, 0xb7b7d2d3, 0x2d2d4848, 0xd2d2b7b8, 0x3635edee, 0xc9ca1212, 0xedee3636, - 0x1211c9ca, 0x48481212, 0xb7b7edee, 0x12124848, 0xededb7b8, 0x48484848, 0xb7b7b7b8, 0x51510000, - 0xaeaf0000, 0x00005151, 0xffffaeaf, 0x3635c9ca, 0xc9ca3636, 0x4847dbdc, 0xb7b82424, 0xdbdc4848, - 0x2423b7b8, 0x75752d2d, 0x8a8ad2d3, 0x2d2d7575, 0xd2d28a8b, 0x7e7e5151, 0x8181aeaf, 0x51517e7e, - 0xaeae8182, 0x6362e4e5, 0x9c9d1b1b, 0xe4e56363, 0x1b1a9c9d, 0x6362aeaf, 0x9c9d5151, 0xaeaf6363, - 0x51509c9d, 0x36363636, 0xc9c9c9ca, 0x6c6c6c6c, 0x93939394, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, - 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0xf6f6f6f7, - 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, - 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0x09090000, 0x09090000, 0x09090000, 0x09090000, - 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, - 0x09090000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, - 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0x00000909, 0x00000909, - 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, - 0x00000909, 0x00000909, 0x00000909, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, - 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, - 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, - 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0xedededee, 0xedededee, 0xedededee, - 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, - 0xedededee, 0xedededee, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, - 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0xe4e4f6f7, - 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, - 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, - 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, - 0x09091b1b, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, - 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, - 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0504fafb, 0xfafb0505, 0xfafb0505, - 0x0504fafb, 0x0b0b0606, 0xf4f4f9fa, 0x06060b0b, 0xf9f9f4f5, 0x08080000, 0xf7f80000, 0x00000808, - 0xfffff7f8, 0x0b0b0b0b, 0xf4f4f4f5, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x11110c0c, - 0xeeeef3f4, 0x0c0c1111, 0xf3f3eeef, 0x11111111, 0xeeeeeeef, 0x12120606, 0xededf9fa, 0x06061212, - 0xf9f9edee, 0x0b0af7f8, 0xf4f50808, 0xf7f80b0b, 0x0807f4f5, 0x0f0f0000, 0xf0f10000, 0x00000f0f, - 0xfffff0f1, 0x14140000, 0xebec0000, 0x00001414, 0xffffebec, 0x19191212, 0xe6e6edee, 0x12121919, - 0xedede6e7, 0x19190b0b, 0xe6e6f4f5, 0x0b0b1919, 0xf4f4e6e7, 0x19191919, 0xe6e6e6e7, 0x0e0df1f2, - 0xf1f20e0e, 0xf1f20e0e, 0x0e0df1f2, 0x1a1a0000, 0xe5e60000, 0x00001a1a, 0xffffe5e6, 0x1211f4f5, - 0xedee0b0b, 0xf4f51212, 0x0b0aedee, 0x1615f8f9, 0xe9ea0707, 0xf8f91616, 0x0706e9ea, 0x22221a1a, - 0xdddde5e6, 0x1a1a2222, 0xe5e5ddde, 0x22221212, 0xddddedee, 0x12122222, 0xededddde, 0x22222222, - 0xddddddde, 0x23230b0b, 0xdcdcf4f5, 0x0b0b2323, 0xf4f4dcdd, 0x1d1d0000, 0xe2e30000, 0x00001d1d, - 0xffffe2e3, 0x1615eced, 0xe9ea1313, 0xeced1616, 0x1312e9ea, 0x1a19f0f1, 0xe5e60f0f, 0xf0f11a1a, - 0x0f0ee5e6, 0x25250000, 0xdadb0000, 0x00002525, 0xffffdadb, 0x2c2c1b1b, 0xd3d3e4e5, 0x1b1b2c2c, - 0xe4e4d3d4, 0x2c2c2424, 0xd3d3dbdc, 0x24242c2c, 0xdbdbd3d4, 0x2c2c1212, 0xd3d3edee, 0x12122c2c, - 0xededd3d4, 0x2120f5f6, 0xdedf0a0a, 0xf5f62121, 0x0a09dedf, 0x2d2d2d2d, 0xd2d2d2d3, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, - 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, - 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, - 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, - 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, - 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, - 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, - 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, - 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, - 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, - 0xf8f90000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0403fbfc, 0xfbfc0404, 0xf9fa0a0a, - 0x0605f5f6, 0xf3f40000, 0x0c0c0000, 0xf3f3f9fa, 0xf3f40606, 0x0c0bf9fa, 0x0c0c0606, 0xfffff1f2, - 0x00000e0e, 0x0c0c0c0c, 0xf3f3f3f4, 0xedee0000, 0x12120000, 0xf3f40e0e, 0x0c0bf1f2, 0xf9f9edee, - 0xf9fa1212, 0x0605edee, 0x06061212, 0xededf5f6, 0xedee0a0a, 0x1211f5f6, 0x12120a0a, 0xffffe9ea, - 0x00001616, 0xe7e80000, 0x18180000, 0xf3f3e9ea, 0xf3f41616, 0x0c0be9ea, 0x0c0c1616, 0xe7e7f7f8, - 0xe7e80808, 0x1817f7f8, 0x18180808, 0xf9f9e5e6, 0xf9fa1a1a, 0x0605e5e6, 0x06061a1a, 0xffffe3e4, - 0x00001c1c, 0x14141414, 0xebebebec, 0xe5e5f1f2, 0x1a1a0e0e, 0xf3f3e1e2, 0x0c0c1e1e, 0xdfdff5f6, - 0x20200a0a, 0xdfdfedee, 0x20201212, 0xe5e5e5e6, 0x1a1a1a1a, 0xebebddde, 0x14142222, 0xf3f3d9da, - 0x0c0c2626, 0xdfdfdfe0, 0x20202020, 0x20202020, 0xd7d7e9ea, 0xddddddde, 0x22222222, 0x00000000, - 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, - 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, - 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, - 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, - 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, - 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, - 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, - 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, - 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, - 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, - 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, - 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, - 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x0605f9fa, 0xf9fa0606, 0xf7f80e0e, - 0x0807f1f2, 0xffffedee, 0x00001212, 0xeff00a0a, 0x100ff5f6, 0xe7e80000, 0x18180000, 0xf7f7e7e8, - 0xf7f81818, 0x0807e7e8, 0x08081818, 0x12121212, 0xedededee, 0xeff01414, 0x100febec, 0xe5e5f1f2, - 0xe5e60e0e, 0x1a19f1f2, 0x1a1a0e0e, 0xffffe1e2, 0x00001e1e, 0xddde0000, 0x22220000, 0xf7f7ddde, - 0xf7f82222, 0x0807ddde, 0x08082222, 0xedede1e2, 0xedee1e1e, 0x1211e1e2, 0x12121e1e, 0xddddf5f6, - 0xddde0a0a, 0x2221f5f6, 0x22220a0a, 0xddddebec, 0x22221414, 0xffffd7d8, 0x00002828, 0x1e1e1e1e, - 0xe1e1e1e2, 0xededd7d8, 0x12122828, 0xd3d40000, 0x2c2c0000, 0xd3d3eff0, 0x2c2c1010, 0xdbdbdbdc, - 0xdbdbdbdc, 0x24242424, 0xd3d3e5e6, 0x2c2c1a1a, 0xe5e5d1d2, 0x1a1a2e2e, 0xededcbcc, 0x12123434, - 0xc9c9ebec, 0xd3d3d3d4, 0x2c2c2c2c, 0xc9c9dfe0, 0xd1d1d1d2, 0xd1d1d1d2, 0x2e2e2e2e, 0x00000000, - 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, - 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, - 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, - 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, - 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, - 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, - 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, - 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, - 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, - 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, - 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, - 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, - 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, - 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, - 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, - 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, - 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, - 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, - 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, - 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, - 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, - 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x0807f7f8, 0xf7f80808, 0xeff00808, - 0x100ff7f8, 0xe7e80000, 0x18180000, 0xf7f7e7e8, 0xf7f81818, 0x0807e7e8, 0x08081818, 0xeff01414, - 0x100febec, 0xffffe3e4, 0x00001c1c, 0xe7e7eff0, 0xe7e81010, 0x1817eff0, 0x18181010, 0xdfe00000, - 0x20200000, 0xefefe3e4, 0xeff01c1c, 0x100fe3e4, 0x10101c1c, 0xdfdff7f8, 0xdfe00808, 0xf7f7dfe0, - 0xf7f82020, 0x0807dfe0, 0x08082020, 0x201ff7f8, 0x20200808, 0x18181818, 0xe7e7e7e8, 0xe7e81818, - 0x1817e7e8, 0xdfdfebec, 0x20201414, 0xffffd7d8, 0x00002828, 0xefefd7d8, 0x10102828, 0xd3d40000, - 0xd3d40000, 0xffffd3d4, 0x00002c2c, 0x2c2c0000, 0x2c2c0000, 0xdfdfdfe0, 0x20202020, 0xd3d3eff0, - 0x2c2c1010, 0xd3d3e7e8, 0xe7e7d3d4, 0x18182c2c, 0x2c2c1818, 0xefefcfd0, 0x10103030, 0xdbdbdbdc, - 0xdbdbdbdc, 0x24242424, 0x24242424, 0xcbcbebec, 0x28282828, 0xd7d7d7d8, 0xcbcbdfe0, 0x00000000, - 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, - 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, - 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, - 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, - 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, - 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, - 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, - 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, - 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, - 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, - 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, - 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, - 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, - 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, - 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, - 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, - 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, - 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, - 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, - 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, - 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, - 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, - 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, - 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, - 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, - 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, - 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, - 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, - 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, - 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, - 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, - 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, - 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x04040404, + 0xfbfbfbfc, 0x05050101, 0xfafafeff, 0x01010505, 0xfefefafb, 0x0403fbfc, 0xfbfc0404, 0x0605fdfe, + 0xf9fa0202, 0xfdfe0606, 0x0201f9fa, 0x09090404, 0xf6f6fbfc, 0x04040909, 0xfbfbf6f7, 0x09090909, + 0xf6f6f6f7, 0x0a0a0101, 0xf5f5feff, 0x01010a0a, 0xfefef5f6, 0x0807fafb, 0xf7f80505, 0xfafb0808, + 0x0504f7f8, 0x0f0f0909, 0xf0f0f6f7, 0x09090f0f, 0xf6f6f0f1, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, + 0x0302f3f4, 0x10100404, 0xefeffbfc, 0x04041010, 0xfbfbeff0, 0x10101010, 0xefefeff0, 0x12120000, + 0xedee0000, 0x00001212, 0xffffedee, 0x0c0bf3f4, 0xf3f40c0c, 0x100ff6f7, 0xeff00909, 0xf6f71010, + 0x0908eff0, 0x1b1b0b0b, 0xe4e4f4f5, 0x0b0b1b1b, 0xf4f4e4e5, 0x1c1c1313, 0xe3e3eced, 0x13131c1c, + 0xecece3e4, 0x1615f9fa, 0xe9ea0606, 0xf9fa1616, 0x0605e9ea, 0x1d1d0404, 0xe2e2fbfc, 0x04041d1d, + 0xfbfbe2e3, 0x1e1e1e1e, 0xe1e1e1e2, 0x2120fdfe, 0xdedf0202, 0xfdfe2121, 0x0201dedf, 0x1716edee, + 0xe8e91212, 0xedee1717, 0x1211e8e9, 0x1e1df0f1, 0xe1e20f0f, 0xf0f11e1e, 0x0f0ee1e2, 0x2e2e1616, + 0xd1d1e9ea, 0x16162e2e, 0xe9e9d1d2, 0x2f2f0d0d, 0xd0d0f2f3, 0x0d0d2f2f, 0xf2f2d0d1, 0x31312323, + 0xcecedcdd, 0x23233131, 0xdcdccecf, 0x2928f4f5, 0xd6d70b0b, 0xf4f52929, 0x0b0ad6d7, 0x33330404, + 0xccccfbfc, 0x04043333, 0xfbfbcccd, 0x36363636, 0xc9c9c9ca, 0x2221ddde, 0xddde2222, 0x2a29e2e3, + 0xd5d61d1d, 0xe2e32a2a, 0x1d1cd5d6, 0x3c3bf9fa, 0xc3c40606, 0xf9fa3c3c, 0x0605c3c4, 0x4c4c1b1b, + 0xb3b3e4e5, 0x1b1b4c4c, 0xe4e4b3b4, 0x4d4d2b2b, 0xb2b2d4d5, 0x2b2b4d4d, 0xd4d4b2b3, 0x3736e7e8, + 0xc8c91818, 0xe7e83737, 0x1817c8c9, 0x4f4f0e0e, 0xb0b0f1f2, 0x0e0e4f4f, 0xf1f1b0b1, 0x53533f3f, + 0xacacc0c1, 0x3f3f5353, 0xc0c0acad, 0x4a49ebec, 0xb5b61414, 0xebec4a4a, 0x1413b5b6, 0x58580202, + 0xa7a7fdfe, 0x02025858, 0xfdfda7a8, 0x5d5d5d5d, 0xa2a2a2a3, 0x3d3ccbcc, 0xc2c33434, 0xcbcc3d3d, + 0x3433c2c3, 0x78783434, 0x8787cbcc, 0x34347878, 0xcbcb8788, 0x4b4ad2d3, 0xb4b52d2d, 0xd2d34b4b, + 0x2d2cb4b5, 0x7d7d4b4b, 0x8282b4b5, 0x4b4b7d7d, 0xb4b48283, 0x7a7a2121, 0x8585dedf, 0x21217a7a, + 0xdede8586, 0x6766f2f3, 0x98990d0d, 0xf2f36767, 0x0d0c9899, 0x605fd7d8, 0x9fa02828, 0xd7d86060, + 0x28279fa0, 0x7f7eddde, 0x80812222, 0xddde7f7f, 0x22218081, 0x5958a6a7, 0xa6a75959, 0x6968b1b2, + 0x96974e4e, 0xb1b26969, 0x4e4d9697, 0x0c0c0c0c, 0xf3f3f3f4, 0x17171717, 0xe8e8e8e9, 0x2a2a2a2a, + 0xd5d5d5d6, 0x49494949, 0xb6b6b6b7, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0x0302feff, 0xfcfd0101, + 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfcfd0101, 0xfeff0303, 0xfeff0303, + 0xfeff0303, 0xfeff0303, 0xfeff0303, 0xfeff0303, 0xfeff0303, 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, + 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, 0x0100fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, + 0xf8f8f8f9, 0x08080202, 0xf7f7fdfe, 0x02020808, 0xfdfdf7f8, 0x0908fdfe, 0xf6f70202, 0xfdfe0909, + 0x0201f6f7, 0x0605f9fa, 0xf9fa0606, 0x0d0d0606, 0xf2f2f9fa, 0x06060d0d, 0xf9f9f2f3, 0x0d0d0d0d, + 0xf2f2f2f3, 0x0e0e0101, 0xf1f1feff, 0x01010e0e, 0xfefef1f2, 0x0c0bf7f8, 0xf3f40808, 0xf7f80c0c, + 0x0807f3f4, 0x17170e0e, 0xe8e8f1f2, 0x0e0e1717, 0xf1f1e8e9, 0x1211fafb, 0xedee0505, 0xfafb1212, + 0x0504edee, 0x18180606, 0xe7e7f9fa, 0x06061818, 0xf9f9e7e8, 0x18181818, 0xe7e7e7e8, 0x1b1afeff, + 0xe4e50101, 0xfeff1b1b, 0x0100e4e5, 0x1110eeef, 0xeeef1111, 0x1716f2f3, 0xe8e90d0d, 0xf2f31717, + 0x0d0ce8e9, 0x28281010, 0xd7d7eff0, 0x10102828, 0xefefd7d8, 0x29291c1c, 0xd6d6e3e4, 0x1c1c2929, + 0xe3e3d6d7, 0x2120f6f7, 0xdedf0909, 0xf6f72121, 0x0908dedf, 0x2b2b0606, 0xd4d4f9fa, 0x06062b2b, + 0xf9f9d4d5, 0x2e2e2e2e, 0xd1d1d1d2, 0x3231fbfc, 0xcdce0404, 0xfbfc3232, 0x0403cdce, 0x2221e4e5, + 0xddde1b1b, 0xe4e52222, 0x1b1addde, 0x2d2ce9ea, 0xd2d31616, 0xe9ea2d2d, 0x1615d2d3, 0x45452222, + 0xbabaddde, 0x22224545, 0xddddbabb, 0x46461313, 0xb9b9eced, 0x13134646, 0xececb9ba, 0x49493535, + 0xb6b6cacb, 0x35354949, 0xcacab6b7, 0x3e3deeef, 0xc1c21111, 0xeeef3e3e, 0x1110c1c2, 0x4d4d0505, + 0xb2b2fafb, 0x05054d4d, 0xfafab2b3, 0x52525252, 0xadadadae, 0x3332cccd, 0xcccd3333, 0x403fd4d5, + 0xbfc02b2b, 0xd4d54040, 0x2b2abfc0, 0x5a59f5f6, 0xa5a60a0a, 0xf5f65a5a, 0x0a09a5a6, 0x72722929, + 0x8d8dd6d7, 0x29297272, 0xd6d68d8e, 0x74744040, 0x8b8bbfc0, 0x40407474, 0xbfbf8b8c, 0x5251dadb, + 0xadae2525, 0xdadb5252, 0x2524adae, 0x77771616, 0x8888e9ea, 0x16167777, 0xe9e98889, 0x7c7c5f5f, + 0x8383a0a1, 0x5f5f7c7c, 0xa0a08384, 0x6f6ee1e2, 0x90911e1e, 0xe1e26f6f, 0x1e1d9091, 0x5c5bb1b2, + 0xa3a44e4e, 0xb1b25c5c, 0x4e4da3a4, 0x7170bbbc, 0x8e8f4444, 0xbbbc7171, 0x44438e8f, 0x12121212, + 0xedededee, 0x22222222, 0xddddddde, 0x3f3f3f3f, 0xc0c0c0c1, 0x6d6d6d6d, 0x92929293, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, + 0x03030303, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, + 0xfcfcfcfd, 0xfcfcfcfd, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, 0x0403feff, + 0x0403feff, 0x0403feff, 0x0403feff, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, + 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfbfc0101, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, + 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0xfeff0404, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, + 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x0100fbfc, 0x07070707, 0x07070707, + 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, + 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, + 0xf5f5fcfd, 0x03030a0a, 0xfcfcf5f6, 0x09090909, 0xf6f6f6f7, 0x0706f8f9, 0xf8f90707, 0x0c0bfcfd, + 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x11110808, 0xeeeef7f8, 0x08081111, 0xf7f7eeef, 0x11111111, + 0xeeeeeeef, 0x13130101, 0xececfeff, 0x01011313, 0xfefeeced, 0x100ff4f5, 0xeff00b0b, 0xf4f51010, + 0x0b0aeff0, 0x1716f9fa, 0xe8e90606, 0xf9fa1717, 0x0605e8e9, 0x1f1f1212, 0xe0e0edee, 0x12121f1f, + 0xedede0e1, 0x20200808, 0xdfdff7f8, 0x08082020, 0xf7f7dfe0, 0x21212121, 0xdedededf, 0x2423feff, + 0xdbdc0101, 0xfeff2424, 0x0100dbdc, 0x1716e8e9, 0xe8e91717, 0x1f1eeeef, 0xe0e11111, 0xeeef1f1f, + 0x1110e0e1, 0x36361515, 0xc9c9eaeb, 0x15153636, 0xeaeac9ca, 0x37372525, 0xc8c8dadb, 0x25253737, + 0xdadac8c9, 0x2c2bf3f4, 0xd3d40c0c, 0xf3f42c2c, 0x0c0bd3d4, 0x39390808, 0xc6c6f7f8, 0x08083939, + 0xf7f7c6c7, 0x3d3d3d3d, 0xc2c2c2c3, 0x4241fafb, 0xbdbe0505, 0xfafb4242, 0x0504bdbe, 0x2d2cdbdc, + 0xd2d32424, 0xdbdc2d2d, 0x2423d2d3, 0x3c3be2e3, 0xc3c41d1d, 0xe2e33c3c, 0x1d1cc3c4, 0x5c5c2d2d, + 0xa3a3d2d3, 0x2d2d5c5c, 0xd2d2a3a4, 0x5d5d1919, 0xa2a2e6e7, 0x19195d5d, 0xe6e6a2a3, 0x61614747, + 0x9e9eb8b9, 0x47476161, 0xb8b89e9f, 0x5352e9ea, 0xacad1616, 0xe9ea5353, 0x1615acad, 0x66660707, + 0x9999f8f9, 0x07076666, 0xf8f8999a, 0x6d6d6d6d, 0x92929293, 0x4443bbbc, 0xbbbc4444, 0x5554c6c7, + 0xaaab3939, 0xc6c75555, 0x3938aaab, 0x7877f2f3, 0x87880d0d, 0xf2f37878, 0x0d0c8788, 0x6e6dcecf, + 0x91923131, 0xcecf6e6e, 0x31309192, 0x7b7a9798, 0x84856868, 0x97987b7b, 0x68678485, 0x18181818, + 0xe7e7e7e8, 0x2e2e2e2e, 0xd1d1d1d2, 0x54545454, 0xabababac, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040404, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, + 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, + 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0x0504feff, 0xfafb0101, 0xfafb0101, 0xfafb0101, + 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfafb0101, 0xfeff0505, + 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, 0xfeff0505, + 0xfeff0505, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0100fafb, + 0x0100fafb, 0x0100fafb, 0x0100fafb, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, + 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0x0a0a0303, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, + 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0xf5f5fcfd, 0x03030a0a, + 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, 0x03030a0a, + 0x03030a0a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, + 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x03030d0d, 0xfcfcf2f3, 0x0908f6f7, 0xf6f70909, 0x0f0efbfc, + 0xf0f10404, 0xfbfc0f0f, 0x0403f0f1, 0x16160b0b, 0xe9e9f4f5, 0x0b0b1616, 0xf4f4e9ea, 0x15151515, + 0xeaeaeaeb, 0x18180202, 0xe7e7fdfe, 0x02021818, 0xfdfde7e8, 0x1413f1f2, 0xebec0e0e, 0xf1f21414, + 0x0e0debec, 0x26261717, 0xd9d9e8e9, 0x17172626, 0xe8e8d9da, 0x1d1cf7f8, 0xe2e30808, 0xf7f81d1d, + 0x0807e2e3, 0x27270b0b, 0xd8d8f4f5, 0x0b0b2727, 0xf4f4d8d9, 0x29292929, 0xd6d6d6d7, 0x2d2cfeff, + 0xd2d30101, 0xfeff2d2d, 0x0100d2d3, 0x1d1ce2e3, 0xe2e31d1d, 0x2726e9ea, 0xd8d91616, 0xe9ea2727, + 0x1615d8d9, 0x43431b1b, 0xbcbce4e5, 0x1b1b4343, 0xe4e4bcbd, 0x45452f2f, 0xbabad0d1, 0x2f2f4545, + 0xd0d0babb, 0x3837f0f1, 0xc7c80f0f, 0xf0f13838, 0x0f0ec7c8, 0x47470b0b, 0xb8b8f4f5, 0x0b0b4747, + 0xf4f4b8b9, 0x4c4c4c4c, 0xb3b3b3b4, 0x5352f9fa, 0xacad0606, 0xf9fa5353, 0x0605acad, 0x3938d2d3, + 0xc6c72d2d, 0xd2d33939, 0x2d2cc6c7, 0x4b4adbdc, 0xb4b52424, 0xdbdc4b4b, 0x2423b4b5, 0x73733838, + 0x8c8cc7c8, 0x38387373, 0xc7c78c8d, 0x75751f1f, 0x8a8ae0e1, 0x1f1f7575, 0xe0e08a8b, 0x7a7a5858, + 0x8585a7a8, 0x58587a7a, 0xa7a78586, 0x6867e3e4, 0x97981c1c, 0xe3e46868, 0x1c1b9798, 0x5554aaab, + 0xaaab5555, 0x6a69b7b8, 0x95964848, 0xb7b86a6a, 0x48479596, 0x1e1e1e1e, 0xe1e1e1e2, 0x3a3a3a3a, + 0xc5c5c5c6, 0x69696969, 0x96969697, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x05050505, 0x05050505, + 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, + 0x05050505, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, + 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, + 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0x0706fdfe, 0xf8f90202, + 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, 0xf8f90202, + 0xf8f90202, 0xf8f90202, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, + 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0xfdfe0707, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, + 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, 0x0201f8f9, + 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, + 0x0b0b0b0b, 0x0b0b0b0b, 0x0b0b0b0b, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, + 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0xf4f4f4f5, 0x0d0d0303, 0x0d0d0303, + 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, 0x0d0d0303, + 0x0d0d0303, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, + 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0xf2f2fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, + 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0xfbfbf0f1, 0x0b0af4f5, 0xf4f50b0b, 0x1211fafb, + 0xedee0505, 0xfafb1212, 0x0504edee, 0x1a1a0d0d, 0xe5e5f2f3, 0x0d0d1a1a, 0xf2f2e5e6, 0x1a1a1a1a, + 0xe5e5e5e6, 0x1d1d0202, 0xe2e2fdfe, 0x02021d1d, 0xfdfde2e3, 0x1817eff0, 0xe7e81010, 0xeff01818, + 0x100fe7e8, 0x2e2e1c1c, 0xd1d1e3e4, 0x1c1c2e2e, 0xe3e3d1d2, 0x2322f6f7, 0xdcdd0909, 0xf6f72323, + 0x0908dcdd, 0x2f2f0d0d, 0xd0d0f2f3, 0x0d0d2f2f, 0xf2f2d0d1, 0x31313131, 0xcecececf, 0x3635feff, + 0xc9ca0101, 0xfeff3636, 0x0100c9ca, 0x2322dcdd, 0xdcdd2323, 0x2f2ee5e6, 0xd0d11a1a, 0xe5e62f2f, + 0x1a19d0d1, 0x51512020, 0xaeaedfe0, 0x20205151, 0xdfdfaeaf, 0x53533838, 0xacacc7c8, 0x38385353, + 0xc7c7acad, 0x4342edee, 0xbcbd1212, 0xedee4343, 0x1211bcbd, 0x56560d0d, 0xa9a9f2f3, 0x0d0d5656, + 0xf2f2a9aa, 0x5b5b5b5b, 0xa4a4a4a5, 0x6362f8f9, 0x9c9d0707, 0xf8f96363, 0x07069c9d, 0x4443c9ca, + 0xbbbc3636, 0xc9ca4444, 0x3635bbbc, 0x5a59d3d4, 0xa5a62c2c, 0xd3d45a5a, 0x2c2ba5a6, 0x7c7bdedf, + 0x83842121, 0xdedf7c7c, 0x21208384, 0x67669899, 0x98996767, 0x7f7ea9aa, 0x80815656, 0xa9aa7f7f, + 0x56558081, 0x25252525, 0xdadadadb, 0x45454545, 0xbabababb, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, + 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0x0807fdfe, 0xf7f80202, 0xf7f80202, 0xf7f80202, + 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, 0xf7f80202, + 0xf7f80202, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, + 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0xfdfe0808, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, + 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, 0x0201f7f8, + 0x0201f7f8, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, + 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0x0d0d0d0d, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, + 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, 0xf2f2f2f3, + 0xf2f2f2f3, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, + 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0x0f0f0404, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, + 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, 0xf0f0fbfc, + 0xf0f0fbfc, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, + 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x04040f0f, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, + 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0xfafaedee, 0x0d0cf2f3, 0xf2f30d0d, 0x1514f9fa, + 0xeaeb0606, 0xf9fa1515, 0x0605eaeb, 0x1e1e0f0f, 0xe1e1f0f1, 0x0f0f1e1e, 0xf0f0e1e2, 0x1e1e1e1e, + 0xe1e1e1e2, 0x22220202, 0xddddfdfe, 0x02022222, 0xfdfdddde, 0x1c1beced, 0xe3e41313, 0xeced1c1c, + 0x1312e3e4, 0x36362020, 0xc9c9dfe0, 0x20203636, 0xdfdfc9ca, 0x2928f4f5, 0xd6d70b0b, 0xf4f52929, + 0x0b0ad6d7, 0x37370f0f, 0xc8c8f0f1, 0x0f0f3737, 0xf0f0c8c9, 0x39393939, 0xc6c6c6c7, 0x3f3efeff, + 0xc0c10101, 0xfeff3f3f, 0x0100c0c1, 0x2827d7d8, 0xd7d82828, 0x3736e1e2, 0xc8c91e1e, 0xe1e23737, + 0x1e1dc8c9, 0x5e5e2525, 0xa1a1dadb, 0x25255e5e, 0xdadaa1a2, 0x60604141, 0x9f9fbebf, 0x41416060, + 0xbebe9fa0, 0x4e4deaeb, 0xb1b21515, 0xeaeb4e4e, 0x1514b1b2, 0x64640f0f, 0x9b9bf0f1, 0x0f0f6464, + 0xf0f09b9c, 0x6a6a6a6a, 0x95959596, 0x7473f7f8, 0x8b8c0808, 0xf7f87474, 0x08078b8c, 0x4f4ec0c1, + 0xb0b13f3f, 0xc0c14f4f, 0x3f3eb0b1, 0x6968cccd, 0x96973333, 0xcccd6969, 0x33329697, 0x78778788, + 0x87887878, 0x2b2b2b2b, 0xd4d4d4d5, 0x50505050, 0xafafafb0, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, + 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, + 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, + 0xf8f8f8f9, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, + 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0x0a09fcfd, 0xf5f60303, 0xf5f60303, 0xf5f60303, + 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, 0xf5f60303, + 0xf5f60303, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, + 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0xfcfd0a0a, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, + 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, 0x0302f5f6, + 0x0302f5f6, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, + 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0xefefeff0, 0xefefeff0, 0xefefeff0, + 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, + 0xefefeff0, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, + 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0x12120505, 0xededfafb, 0xededfafb, 0xededfafb, + 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, 0xededfafb, + 0xededfafb, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, + 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x05051212, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, + 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0xfafaebec, 0x0f0ef0f1, 0xf0f10f0f, 0x1817f8f9, + 0xe7e80707, 0xf8f91818, 0x0706e7e8, 0x23231111, 0xdcdceeef, 0x11112323, 0xeeeedcdd, 0x22222222, + 0xddddddde, 0x26260303, 0xd9d9fcfd, 0x03032626, 0xfcfcd9da, 0x201fe9ea, 0xdfe01616, 0xe9ea2020, + 0x1615dfe0, 0x3d3d2525, 0xc2c2dadb, 0x25253d3d, 0xdadac2c3, 0x2f2ef2f3, 0xd0d10d0d, 0xf2f32f2f, + 0x0d0cd0d1, 0x3f3f1111, 0xc0c0eeef, 0x11113f3f, 0xeeeec0c1, 0x41414141, 0xbebebebf, 0x4847feff, + 0xb7b80101, 0xfeff4848, 0x0100b7b8, 0x2e2dd1d2, 0xd1d22e2e, 0x3f3edcdd, 0xc0c12323, 0xdcdd3f3f, + 0x2322c0c1, 0x6b6b2b2b, 0x9494d4d5, 0x2b2b6b6b, 0xd4d49495, 0x6e6e4b4b, 0x9191b4b5, 0x4b4b6e6e, + 0xb4b49192, 0x5958e7e8, 0xa6a71818, 0xe7e85959, 0x1817a6a7, 0x72721111, 0x8d8deeef, 0x11117272, + 0xeeee8d8e, 0x79797979, 0x86868687, 0x5b5ab7b8, 0xa4a54848, 0xb7b85b5b, 0x4847a4a5, 0x7877c5c6, + 0x87883a3a, 0xc5c67878, 0x3a398788, 0x31313131, 0xcecececf, 0x5c5c5c5c, 0xa3a3a3a4, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, + 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0xf7f7f7f8, + 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, + 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, + 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0x0b0afcfd, 0xf4f50303, + 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xf4f50303, + 0xf4f50303, 0xf4f50303, 0xf4f50303, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, + 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0xfcfd0b0b, 0x0302f4f5, + 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, + 0x0302f4f5, 0x0302f4f5, 0x0302f4f5, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, + 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0xedededee, + 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, + 0xedededee, 0xedededee, 0xedededee, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, + 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0x14140505, 0xebebfafb, + 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, 0xebebfafb, + 0xebebfafb, 0xebebfafb, 0xebebfafb, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, + 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x05051414, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, + 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x1110eeef, 0xeeef1111, 0x1b1af8f9, + 0xe4e50707, 0xf8f91b1b, 0x0706e4e5, 0x27271313, 0xd8d8eced, 0x13132727, 0xececd8d9, 0x27272727, + 0xd8d8d8d9, 0x2b2b0303, 0xd4d4fcfd, 0x03032b2b, 0xfcfcd4d5, 0x2423e7e8, 0xdbdc1818, 0xe7e82424, + 0x1817dbdc, 0x45452a2a, 0xbabad5d6, 0x2a2a4545, 0xd5d5babb, 0x3534f1f2, 0xcacb0e0e, 0xf1f23535, + 0x0e0dcacb, 0x47471313, 0xb8b8eced, 0x13134747, 0xececb8b9, 0x49494949, 0xb6b6b6b7, 0x504ffdfe, + 0xafb00202, 0xfdfe5050, 0x0201afb0, 0x3433cbcc, 0xcbcc3434, 0x4645d8d9, 0xb9ba2727, 0xd8d94646, + 0x2726b9ba, 0x79793030, 0x8686cfd0, 0x30307979, 0xcfcf8687, 0x7c7c5454, 0x8383abac, 0x54547c7c, + 0xabab8384, 0x6463e4e5, 0x9b9c1b1b, 0xe4e56464, 0x1b1a9b9c, 0x6665aeaf, 0x999a5151, 0xaeaf6666, + 0x5150999a, 0x37373737, 0xc8c8c8c9, 0x68686868, 0x97979798, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, + 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0xf6f6f6f7, + 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, + 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, + 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, 0x0c0bfcfd, + 0x0c0bfcfd, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, + 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xf3f40303, 0xfcfd0c0c, 0xfcfd0c0c, + 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, + 0xfcfd0c0c, 0xfcfd0c0c, 0xfcfd0c0c, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, + 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, 0x0302f3f4, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, + 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0x17170606, 0xe8e8f9fa, + 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, + 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0xe8e8f9fa, 0x06061717, 0x06061717, 0x06061717, 0x06061717, + 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, 0x06061717, + 0x06061717, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, + 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0xf9f9e8e9, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, + 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x0403fbfc, 0xfbfc0404, 0x0605fdfe, + 0xf9fa0202, 0xfdfe0606, 0x0201f9fa, 0x08080404, 0xf7f7fbfc, 0x04040808, 0xfbfbf7f8, 0x08080808, + 0xf7f7f7f8, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x0807fbfc, 0xf7f80404, 0xfbfc0808, + 0x0403f7f8, 0x0e0e0808, 0xf1f1f7f8, 0x08080e0e, 0xf7f7f1f2, 0x0c0bfdfe, 0xf3f40202, 0xfdfe0c0c, + 0x0201f3f4, 0x10100404, 0xefeffbfc, 0x04041010, 0xfbfbeff0, 0x10101010, 0xefefeff0, 0x12120000, + 0xedee0000, 0x00001212, 0xffffedee, 0x0c0bf3f4, 0xf3f40c0c, 0x100ff7f8, 0xeff00808, 0xf7f81010, + 0x0807eff0, 0x1a1a0a0a, 0xe5e5f5f6, 0x0a0a1a1a, 0xf5f5e5e6, 0x1c1c1212, 0xe3e3edee, 0x12121c1c, + 0xedede3e4, 0x1615f9fa, 0xe9ea0606, 0xf9fa1616, 0x0605e9ea, 0x1c1c0404, 0xe3e3fbfc, 0x04041c1c, + 0xfbfbe3e4, 0x1e1e1e1e, 0xe1e1e1e2, 0x201ffdfe, 0xdfe00202, 0xfdfe2020, 0x0201dfe0, 0x1615edee, + 0xe9ea1212, 0xedee1616, 0x1211e9ea, 0x1e1df1f2, 0xe1e20e0e, 0xf1f21e1e, 0x0e0de1e2, 0x2e2e1616, + 0xd1d1e9ea, 0x16162e2e, 0xe9e9d1d2, 0x2e2e0c0c, 0xd1d1f3f4, 0x0c0c2e2e, 0xf3f3d1d2, 0x30302222, + 0xcfcfddde, 0x22223030, 0xddddcfd0, 0x2827f5f6, 0xd7d80a0a, 0xf5f62828, 0x0a09d7d8, 0x32320404, + 0xcdcdfbfc, 0x04043232, 0xfbfbcdce, 0x36363636, 0xc9c9c9ca, 0x2221ddde, 0xddde2222, 0x2a29e3e4, + 0xd5d61c1c, 0xe3e42a2a, 0x1c1bd5d6, 0x3c3bf9fa, 0xc3c40606, 0xf9fa3c3c, 0x0605c3c4, 0x4c4c1a1a, + 0xb3b3e5e6, 0x1a1a4c4c, 0xe5e5b3b4, 0x4c4c2a2a, 0xb3b3d5d6, 0x2a2a4c4c, 0xd5d5b3b4, 0x3635e7e8, + 0xc9ca1818, 0xe7e83636, 0x1817c9ca, 0x4e4e0e0e, 0xb1b1f1f2, 0x0e0e4e4e, 0xf1f1b1b2, 0x52523e3e, + 0xadadc1c2, 0x3e3e5252, 0xc1c1adae, 0x4a49ebec, 0xb5b61414, 0xebec4a4a, 0x1413b5b6, 0x58580202, + 0xa7a7fdfe, 0x02025858, 0xfdfda7a8, 0x5c5c5c5c, 0xa3a3a3a4, 0x3c3bcbcc, 0xc3c43434, 0xcbcc3c3c, + 0x3433c3c4, 0x76763434, 0x8989cbcc, 0x34347676, 0xcbcb898a, 0x4a49d3d4, 0xb5b62c2c, 0xd3d44a4a, + 0x2c2bb5b6, 0x76764a4a, 0x8989b5b6, 0x4a4a7676, 0xb5b5898a, 0x76762020, 0x8989dfe0, 0x20207676, + 0xdfdf898a, 0x6665f3f4, 0x999a0c0c, 0xf3f46666, 0x0c0b999a, 0x605fd7d8, 0x9fa02828, 0xd7d86060, + 0x28279fa0, 0x7675ddde, 0x898a2222, 0xddde7676, 0x2221898a, 0x5857a7a8, 0xa7a85858, 0x6867b1b2, + 0x97984e4e, 0xb1b26868, 0x4e4d9798, 0x0c0c0c0c, 0xf3f3f3f4, 0x16161616, 0xe9e9e9ea, 0x2a2a2a2a, + 0xd5d5d5d6, 0x48484848, 0xb7b7b7b8, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0xfdfe0000, + 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, + 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, + 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, + 0xf9f9f9fa, 0x09090303, 0xf6f6fcfd, 0x03030909, 0xfcfcf6f7, 0x0908fcfd, 0xf6f70303, 0xfcfd0909, + 0x0302f6f7, 0x0605f9fa, 0xf9fa0606, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0xf9f9f3f4, 0x0c0c0c0c, + 0xf3f3f3f4, 0x0f0f0000, 0xf0f10000, 0x00000f0f, 0xfffff0f1, 0x0c0bf6f7, 0xf3f40909, 0xf6f70c0c, + 0x0908f3f4, 0x18180f0f, 0xe7e7f0f1, 0x0f0f1818, 0xf0f0e7e8, 0x1211f9fa, 0xedee0606, 0xf9fa1212, + 0x0605edee, 0x18180606, 0xe7e7f9fa, 0x06061818, 0xf9f9e7e8, 0x18181818, 0xe7e7e7e8, 0x1b1b0000, + 0xe4e50000, 0x00001b1b, 0xffffe4e5, 0x1211edee, 0xedee1212, 0x1817f3f4, 0xe7e80c0c, 0xf3f41818, + 0x0c0be7e8, 0x27270f0f, 0xd8d8f0f1, 0x0f0f2727, 0xf0f0d8d9, 0x2a2a1b1b, 0xd5d5e4e5, 0x1b1b2a2a, + 0xe4e4d5d6, 0x2120f6f7, 0xdedf0909, 0xf6f72121, 0x0908dedf, 0x2a2a0606, 0xd5d5f9fa, 0x06062a2a, + 0xf9f9d5d6, 0x2d2d2d2d, 0xd2d2d2d3, 0x3332fcfd, 0xcccd0303, 0xfcfd3333, 0x0302cccd, 0x2120e4e5, + 0xdedf1b1b, 0xe4e52121, 0x1b1adedf, 0x2d2ceaeb, 0xd2d31515, 0xeaeb2d2d, 0x1514d2d3, 0x45452121, + 0xbabadedf, 0x21214545, 0xdedebabb, 0x45451212, 0xbabaedee, 0x12124545, 0xededbabb, 0x48483636, + 0xb7b7c9ca, 0x36364848, 0xc9c9b7b8, 0x3f3eedee, 0xc0c11212, 0xedee3f3f, 0x1211c0c1, 0x4e4e0606, + 0xb1b1f9fa, 0x06064e4e, 0xf9f9b1b2, 0x51515151, 0xaeaeaeaf, 0x3332cccd, 0xcccd3333, 0x3f3ed5d6, + 0xc0c12a2a, 0xd5d63f3f, 0x2a29c0c1, 0x5a59f6f7, 0xa5a60909, 0xf6f75a5a, 0x0908a5a6, 0x72722a2a, + 0x8d8dd5d6, 0x2a2a7272, 0xd5d58d8e, 0x75753f3f, 0x8a8ac0c1, 0x3f3f7575, 0xc0c08a8b, 0x5150dbdc, + 0xaeaf2424, 0xdbdc5151, 0x2423aeaf, 0x78781515, 0x8787eaeb, 0x15157878, 0xeaea8788, 0x7b7b6060, + 0x84849fa0, 0x60607b7b, 0x9f9f8485, 0x6f6ee1e2, 0x90911e1e, 0xe1e26f6f, 0x1e1d9091, 0x5d5cb1b2, + 0xa2a34e4e, 0xb1b25d5d, 0x4e4da2a3, 0x7271babb, 0x8d8e4545, 0xbabb7272, 0x45448d8e, 0x12121212, + 0xedededee, 0x21212121, 0xdedededf, 0x3f3f3f3f, 0xc0c0c0c1, 0x6c6c6c6c, 0x93939394, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, 0x03030303, + 0x03030303, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, 0xfcfcfcfd, + 0xfcfcfcfd, 0xfcfcfcfd, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, + 0x03030000, 0x03030000, 0x03030000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, + 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0x00000303, 0x00000303, 0x00000303, 0x00000303, + 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, + 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, + 0xf7f7fbfc, 0x04040808, 0xfbfbf7f8, 0x08080808, 0xf7f7f7f8, 0x0807f7f8, 0xf7f80808, 0x0c0bfbfc, + 0xf3f40404, 0xfbfc0c0c, 0x0403f3f4, 0x10100808, 0xefeff7f8, 0x08081010, 0xf7f7eff0, 0x10101010, + 0xefefeff0, 0x14140000, 0xebec0000, 0x00001414, 0xffffebec, 0x100ff3f4, 0xeff00c0c, 0xf3f41010, + 0x0c0beff0, 0x1817fbfc, 0xe7e80404, 0xfbfc1818, 0x0403e7e8, 0x20201010, 0xdfdfeff0, 0x10102020, + 0xefefdfe0, 0x20200808, 0xdfdff7f8, 0x08082020, 0xf7f7dfe0, 0x20202020, 0xdfdfdfe0, 0x24240000, + 0xdbdc0000, 0x00002424, 0xffffdbdc, 0x1817e7e8, 0xe7e81818, 0x201feff0, 0xdfe01010, 0xeff02020, + 0x100fdfe0, 0x34341414, 0xcbcbebec, 0x14143434, 0xebebcbcc, 0x38382424, 0xc7c7dbdc, 0x24243838, + 0xdbdbc7c8, 0x2c2bf3f4, 0xd3d40c0c, 0xf3f42c2c, 0x0c0bd3d4, 0x38380808, 0xc7c7f7f8, 0x08083838, + 0xf7f7c7c8, 0x3c3c3c3c, 0xc3c3c3c4, 0x403ffbfc, 0xbfc00404, 0xfbfc4040, 0x0403bfc0, 0x2c2bdbdc, + 0xd3d42424, 0xdbdc2c2c, 0x2423d3d4, 0x3c3be3e4, 0xc3c41c1c, 0xe3e43c3c, 0x1c1bc3c4, 0x5c5c2c2c, + 0xa3a3d3d4, 0x2c2c5c5c, 0xd3d3a3a4, 0x5c5c1818, 0xa3a3e7e8, 0x18185c5c, 0xe7e7a3a4, 0x60604848, + 0x9f9fb7b8, 0x48486060, 0xb7b79fa0, 0x5453ebec, 0xabac1414, 0xebec5454, 0x1413abac, 0x64640808, + 0x9b9bf7f8, 0x08086464, 0xf7f79b9c, 0x6c6c6c6c, 0x93939394, 0x4443bbbc, 0xbbbc4444, 0x5453c7c8, + 0xabac3838, 0xc7c85454, 0x3837abac, 0x7877f3f4, 0x87880c0c, 0xf3f47878, 0x0c0b8788, 0x6c6bcfd0, + 0x93943030, 0xcfd06c6c, 0x302f9394, 0x7c7b9798, 0x83846868, 0x97987c7c, 0x68678384, 0x18181818, + 0xe7e7e7e8, 0x2c2c2c2c, 0xd3d3d3d4, 0x54545454, 0xabababac, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040404, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, + 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, + 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, + 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0x00000404, + 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, + 0x00000404, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, + 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0x08080404, + 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0x08080404, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, + 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0xf7f7fbfc, 0x04040808, + 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, 0x04040808, + 0x04040808, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, + 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x05050f0f, 0xfafaf0f1, 0x0a09f5f6, 0xf5f60a0a, 0x0f0efafb, + 0xf0f10505, 0xfafb0f0f, 0x0504f0f1, 0x14140a0a, 0xebebf5f6, 0x0a0a1414, 0xf5f5ebec, 0x14141414, + 0xebebebec, 0x19190000, 0xe6e70000, 0x00001919, 0xffffe6e7, 0x1413f0f1, 0xebec0f0f, 0xf0f11414, + 0x0f0eebec, 0x28281919, 0xd7d7e6e7, 0x19192828, 0xe6e6d7d8, 0x1e1df5f6, 0xe1e20a0a, 0xf5f61e1e, + 0x0a09e1e2, 0x28280a0a, 0xd7d7f5f6, 0x0a0a2828, 0xf5f5d7d8, 0x28282828, 0xd7d7d7d8, 0x2d2d0000, + 0xd2d30000, 0x00002d2d, 0xffffd2d3, 0x1e1de1e2, 0xe1e21e1e, 0x2827ebec, 0xd7d81414, 0xebec2828, + 0x1413d7d8, 0x41411919, 0xbebee6e7, 0x19194141, 0xe6e6bebf, 0x46462d2d, 0xb9b9d2d3, 0x2d2d4646, + 0xd2d2b9ba, 0x3736f0f1, 0xc8c90f0f, 0xf0f13737, 0x0f0ec8c9, 0x46460a0a, 0xb9b9f5f6, 0x0a0a4646, + 0xf5f5b9ba, 0x4b4b4b4b, 0xb4b4b4b5, 0x5554fafb, 0xaaab0505, 0xfafb5555, 0x0504aaab, 0x3736d2d3, + 0xc8c92d2d, 0xd2d33737, 0x2d2cc8c9, 0x4b4adcdd, 0xb4b52323, 0xdcdd4b4b, 0x2322b4b5, 0x73733737, + 0x8c8cc8c9, 0x37377373, 0xc8c88c8d, 0x73731e1e, 0x8c8ce1e2, 0x1e1e7373, 0xe1e18c8d, 0x78785a5a, + 0x8787a5a6, 0x5a5a7878, 0xa5a58788, 0x6968e1e2, 0x96971e1e, 0xe1e26969, 0x1e1d9697, 0x5554aaab, + 0xaaab5555, 0x6968b9ba, 0x96974646, 0xb9ba6969, 0x46459697, 0x1e1e1e1e, 0xe1e1e1e2, 0x3c3c3c3c, + 0xc3c3c3c4, 0x69696969, 0x96969697, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x05050505, 0x05050505, + 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, 0x05050505, + 0x05050505, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, + 0xfafafafb, 0xfafafafb, 0xfafafafb, 0xfafafafb, 0x05050000, 0x05050000, 0x05050000, 0x05050000, + 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0x05050000, 0xfafb0000, + 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, 0xfafb0000, + 0xfafb0000, 0xfafb0000, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, + 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0x00000505, 0xfffffafb, 0xfffffafb, 0xfffffafb, + 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, 0xfffffafb, + 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, + 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, + 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0x0f0f0505, 0x0f0f0505, + 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, 0x0f0f0505, + 0x0f0f0505, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, + 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0xf0f0fafb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, + 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0xf9f9f3f4, 0x0c0bf3f4, 0xf3f40c0c, 0x1211f9fa, + 0xedee0606, 0xf9fa1212, 0x0605edee, 0x18180c0c, 0xe7e7f3f4, 0x0c0c1818, 0xf3f3e7e8, 0x18181818, + 0xe7e7e7e8, 0x1e1e0000, 0xe1e20000, 0x00001e1e, 0xffffe1e2, 0x1817edee, 0xe7e81212, 0xedee1818, + 0x1211e7e8, 0x30301e1e, 0xcfcfe1e2, 0x1e1e3030, 0xe1e1cfd0, 0x2423f9fa, 0xdbdc0606, 0xf9fa2424, + 0x0605dbdc, 0x30300c0c, 0xcfcff3f4, 0x0c0c3030, 0xf3f3cfd0, 0x30303030, 0xcfcfcfd0, 0x36360000, + 0xc9ca0000, 0x00003636, 0xffffc9ca, 0x2423dbdc, 0xdbdc2424, 0x302fe7e8, 0xcfd01818, 0xe7e83030, + 0x1817cfd0, 0x4e4e1e1e, 0xb1b1e1e2, 0x1e1e4e4e, 0xe1e1b1b2, 0x54543636, 0xababc9ca, 0x36365454, + 0xc9c9abac, 0x4241edee, 0xbdbe1212, 0xedee4242, 0x1211bdbe, 0x54540c0c, 0xababf3f4, 0x0c0c5454, + 0xf3f3abac, 0x5a5a5a5a, 0xa5a5a5a6, 0x605ff9fa, 0x9fa00606, 0xf9fa6060, 0x06059fa0, 0x4241c9ca, + 0xbdbe3636, 0xc9ca4242, 0x3635bdbe, 0x5a59d5d6, 0xa5a62a2a, 0xd5d65a5a, 0x2a29a5a6, 0x7e7de1e2, + 0x81821e1e, 0xe1e27e7e, 0x1e1d8182, 0x6665999a, 0x999a6666, 0x7e7dabac, 0x81825454, 0xabac7e7e, + 0x54538182, 0x24242424, 0xdbdbdbdc, 0x42424242, 0xbdbdbdbe, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, + 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, + 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, + 0xf9fa0000, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, + 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, + 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, + 0xfffff9fa, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, + 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0x0c0c0606, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, + 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, 0xf3f3f9fa, + 0xf3f3f9fa, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, + 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x06060c0c, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, + 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0xf8f8eaeb, 0x0e0df1f2, 0xf1f20e0e, 0x1514f8f9, + 0xeaeb0707, 0xf8f91515, 0x0706eaeb, 0x1c1c0e0e, 0xe3e3f1f2, 0x0e0e1c1c, 0xf1f1e3e4, 0x1c1c1c1c, + 0xe3e3e3e4, 0x23230000, 0xdcdd0000, 0x00002323, 0xffffdcdd, 0x1c1beaeb, 0xe3e41515, 0xeaeb1c1c, + 0x1514e3e4, 0x38382323, 0xc7c7dcdd, 0x23233838, 0xdcdcc7c8, 0x2a29f1f2, 0xd5d60e0e, 0xf1f22a2a, + 0x0e0dd5d6, 0x38380e0e, 0xc7c7f1f2, 0x0e0e3838, 0xf1f1c7c8, 0x38383838, 0xc7c7c7c8, 0x3f3f0000, + 0xc0c10000, 0x00003f3f, 0xffffc0c1, 0x2a29d5d6, 0xd5d62a2a, 0x3837e3e4, 0xc7c81c1c, 0xe3e43838, + 0x1c1bc7c8, 0x5b5b2323, 0xa4a4dcdd, 0x23235b5b, 0xdcdca4a5, 0x62623f3f, 0x9d9dc0c1, 0x3f3f6262, + 0xc0c09d9e, 0x4d4ceaeb, 0xb2b31515, 0xeaeb4d4d, 0x1514b2b3, 0x62620e0e, 0x9d9df1f2, 0x0e0e6262, + 0xf1f19d9e, 0x69696969, 0x96969697, 0x7776f8f9, 0x88890707, 0xf8f97777, 0x07068889, 0x4d4cc0c1, + 0xb2b33f3f, 0xc0c14d4d, 0x3f3eb2b3, 0x6968cecf, 0x96973131, 0xcecf6969, 0x31309697, 0x77768889, + 0x88897777, 0x2a2a2a2a, 0xd5d5d5d6, 0x4d4d4d4d, 0xb2b2b2b3, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, + 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0x07070707, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, + 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, 0xf8f8f8f9, + 0xf8f8f8f9, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, + 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0xf8f90000, 0xf8f90000, 0xf8f90000, + 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, + 0xf8f90000, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, + 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0x00000707, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, + 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, 0xfffff8f9, + 0xfffff8f9, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, + 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0x0e0e0e0e, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, + 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, 0xf1f1f1f2, + 0xf1f1f1f2, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, + 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0x15150707, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, + 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, 0xeaeaf8f9, + 0xeaeaf8f9, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, + 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x07071515, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, + 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0xf7f7eff0, 0x100feff0, 0xeff01010, 0x1817f7f8, + 0xe7e80808, 0xf7f81818, 0x0807e7e8, 0x20201010, 0xdfdfeff0, 0x10102020, 0xefefdfe0, 0x20202020, + 0xdfdfdfe0, 0x28280000, 0xd7d80000, 0x00002828, 0xffffd7d8, 0x201fe7e8, 0xdfe01818, 0xe7e82020, + 0x1817dfe0, 0x40402828, 0xbfbfd7d8, 0x28284040, 0xd7d7bfc0, 0x302feff0, 0xcfd01010, 0xeff03030, + 0x100fcfd0, 0x40401010, 0xbfbfeff0, 0x10104040, 0xefefbfc0, 0x40404040, 0xbfbfbfc0, 0x48480000, + 0xb7b80000, 0x00004848, 0xffffb7b8, 0x302fcfd0, 0xcfd03030, 0x403fdfe0, 0xbfc02020, 0xdfe04040, + 0x201fbfc0, 0x68682828, 0x9797d7d8, 0x28286868, 0xd7d79798, 0x70704848, 0x8f8fb7b8, 0x48487070, + 0xb7b78f90, 0x5857e7e8, 0xa7a81818, 0xe7e85858, 0x1817a7a8, 0x70701010, 0x8f8feff0, 0x10107070, + 0xefef8f90, 0x78787878, 0x87878788, 0x5857b7b8, 0xa7a84848, 0xb7b85858, 0x4847a7a8, 0x7877c7c8, + 0x87883838, 0xc7c87878, 0x38378788, 0x30303030, 0xcfcfcfd0, 0x58585858, 0xa7a7a7a8, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, + 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0x08080808, 0xf7f7f7f8, + 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, + 0xf7f7f7f8, 0xf7f7f7f8, 0xf7f7f7f8, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, + 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0x08080000, 0xf7f80000, + 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, 0xf7f80000, + 0xf7f80000, 0xf7f80000, 0xf7f80000, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, + 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0x00000808, 0xfffff7f8, + 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, + 0xfffff7f8, 0xfffff7f8, 0xfffff7f8, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, + 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0x10101010, 0xefefeff0, + 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, 0xefefeff0, + 0xefefeff0, 0xefefeff0, 0xefefeff0, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, + 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0x10100808, 0xefeff7f8, + 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, + 0xefeff7f8, 0xefeff7f8, 0xefeff7f8, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, + 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x08081010, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, + 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x1211edee, 0xedee1212, 0x1b1af6f7, + 0xe4e50909, 0xf6f71b1b, 0x0908e4e5, 0x24241212, 0xdbdbedee, 0x12122424, 0xededdbdc, 0x24242424, + 0xdbdbdbdc, 0x2d2d0000, 0xd2d30000, 0x00002d2d, 0xffffd2d3, 0x2423e4e5, 0xdbdc1b1b, 0xe4e52424, + 0x1b1adbdc, 0x48482d2d, 0xb7b7d2d3, 0x2d2d4848, 0xd2d2b7b8, 0x3635edee, 0xc9ca1212, 0xedee3636, + 0x1211c9ca, 0x48481212, 0xb7b7edee, 0x12124848, 0xededb7b8, 0x48484848, 0xb7b7b7b8, 0x51510000, + 0xaeaf0000, 0x00005151, 0xffffaeaf, 0x3635c9ca, 0xc9ca3636, 0x4847dbdc, 0xb7b82424, 0xdbdc4848, + 0x2423b7b8, 0x75752d2d, 0x8a8ad2d3, 0x2d2d7575, 0xd2d28a8b, 0x7e7e5151, 0x8181aeaf, 0x51517e7e, + 0xaeae8182, 0x6362e4e5, 0x9c9d1b1b, 0xe4e56363, 0x1b1a9c9d, 0x6362aeaf, 0x9c9d5151, 0xaeaf6363, + 0x51509c9d, 0x36363636, 0xc9c9c9ca, 0x6c6c6c6c, 0x93939394, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, + 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0x09090909, 0xf6f6f6f7, + 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, + 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0xf6f6f6f7, 0x09090000, 0x09090000, 0x09090000, 0x09090000, + 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, 0x09090000, + 0x09090000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, + 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0xf6f70000, 0x00000909, 0x00000909, + 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, 0x00000909, + 0x00000909, 0x00000909, 0x00000909, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, + 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, 0xfffff6f7, + 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, + 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0x12121212, 0xedededee, 0xedededee, 0xedededee, + 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, 0xedededee, + 0xedededee, 0xedededee, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, + 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0x1b1b0909, 0xe4e4f6f7, + 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, + 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0xe4e4f6f7, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, + 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, 0x09091b1b, + 0x09091b1b, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, + 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0xf6f6e4e5, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, + 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0504fafb, 0xfafb0505, 0xfafb0505, + 0x0504fafb, 0x0b0b0606, 0xf4f4f9fa, 0x06060b0b, 0xf9f9f4f5, 0x08080000, 0xf7f80000, 0x00000808, + 0xfffff7f8, 0x0b0b0b0b, 0xf4f4f4f5, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x11110c0c, + 0xeeeef3f4, 0x0c0c1111, 0xf3f3eeef, 0x11111111, 0xeeeeeeef, 0x12120606, 0xededf9fa, 0x06061212, + 0xf9f9edee, 0x0b0af7f8, 0xf4f50808, 0xf7f80b0b, 0x0807f4f5, 0x0f0f0000, 0xf0f10000, 0x00000f0f, + 0xfffff0f1, 0x14140000, 0xebec0000, 0x00001414, 0xffffebec, 0x19191212, 0xe6e6edee, 0x12121919, + 0xedede6e7, 0x19190b0b, 0xe6e6f4f5, 0x0b0b1919, 0xf4f4e6e7, 0x19191919, 0xe6e6e6e7, 0x0e0df1f2, + 0xf1f20e0e, 0xf1f20e0e, 0x0e0df1f2, 0x1a1a0000, 0xe5e60000, 0x00001a1a, 0xffffe5e6, 0x1211f4f5, + 0xedee0b0b, 0xf4f51212, 0x0b0aedee, 0x1615f8f9, 0xe9ea0707, 0xf8f91616, 0x0706e9ea, 0x22221a1a, + 0xdddde5e6, 0x1a1a2222, 0xe5e5ddde, 0x22221212, 0xddddedee, 0x12122222, 0xededddde, 0x22222222, + 0xddddddde, 0x23230b0b, 0xdcdcf4f5, 0x0b0b2323, 0xf4f4dcdd, 0x1d1d0000, 0xe2e30000, 0x00001d1d, + 0xffffe2e3, 0x1615eced, 0xe9ea1313, 0xeced1616, 0x1312e9ea, 0x1a19f0f1, 0xe5e60f0f, 0xf0f11a1a, + 0x0f0ee5e6, 0x25250000, 0xdadb0000, 0x00002525, 0xffffdadb, 0x2c2c1b1b, 0xd3d3e4e5, 0x1b1b2c2c, + 0xe4e4d3d4, 0x2c2c2424, 0xd3d3dbdc, 0x24242c2c, 0xdbdbd3d4, 0x2c2c1212, 0xd3d3edee, 0x12122c2c, + 0xededd3d4, 0x2120f5f6, 0xdedf0a0a, 0xf5f62121, 0x0a09dedf, 0x2d2d2d2d, 0xd2d2d2d3, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, + 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, + 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, + 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, + 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, + 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, + 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, + 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, + 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x03030000, 0xfcfd0000, + 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, 0xf8f90000, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x07070000, + 0xf8f90000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0403fbfc, 0xfbfc0404, 0xf9fa0a0a, + 0x0605f5f6, 0xf3f40000, 0x0c0c0000, 0xf3f3f9fa, 0xf3f40606, 0x0c0bf9fa, 0x0c0c0606, 0xfffff1f2, + 0x00000e0e, 0x0c0c0c0c, 0xf3f3f3f4, 0xedee0000, 0x12120000, 0xf3f40e0e, 0x0c0bf1f2, 0xf9f9edee, + 0xf9fa1212, 0x0605edee, 0x06061212, 0xededf5f6, 0xedee0a0a, 0x1211f5f6, 0x12120a0a, 0xffffe9ea, + 0x00001616, 0xe7e80000, 0x18180000, 0xf3f3e9ea, 0xf3f41616, 0x0c0be9ea, 0x0c0c1616, 0xe7e7f7f8, + 0xe7e80808, 0x1817f7f8, 0x18180808, 0xf9f9e5e6, 0xf9fa1a1a, 0x0605e5e6, 0x06061a1a, 0xffffe3e4, + 0x00001c1c, 0x14141414, 0xebebebec, 0xe5e5f1f2, 0x1a1a0e0e, 0xf3f3e1e2, 0x0c0c1e1e, 0xdfdff5f6, + 0x20200a0a, 0xdfdfedee, 0x20201212, 0xe5e5e5e6, 0x1a1a1a1a, 0xebebddde, 0x14142222, 0xf3f3d9da, + 0x0c0c2626, 0xdfdfdfe0, 0x20202020, 0x20202020, 0xd7d7e9ea, 0xddddddde, 0x22222222, 0x00000000, + 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, + 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, + 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, + 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, + 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, + 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, + 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, + 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, + 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, + 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, + 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, + 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, + 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x0605f9fa, 0xf9fa0606, 0xf7f80e0e, + 0x0807f1f2, 0xffffedee, 0x00001212, 0xeff00a0a, 0x100ff5f6, 0xe7e80000, 0x18180000, 0xf7f7e7e8, + 0xf7f81818, 0x0807e7e8, 0x08081818, 0x12121212, 0xedededee, 0xeff01414, 0x100febec, 0xe5e5f1f2, + 0xe5e60e0e, 0x1a19f1f2, 0x1a1a0e0e, 0xffffe1e2, 0x00001e1e, 0xddde0000, 0x22220000, 0xf7f7ddde, + 0xf7f82222, 0x0807ddde, 0x08082222, 0xedede1e2, 0xedee1e1e, 0x1211e1e2, 0x12121e1e, 0xddddf5f6, + 0xddde0a0a, 0x2221f5f6, 0x22220a0a, 0xddddebec, 0x22221414, 0xffffd7d8, 0x00002828, 0x1e1e1e1e, + 0xe1e1e1e2, 0xededd7d8, 0x12122828, 0xd3d40000, 0x2c2c0000, 0xd3d3eff0, 0x2c2c1010, 0xdbdbdbdc, + 0xdbdbdbdc, 0x24242424, 0xd3d3e5e6, 0x2c2c1a1a, 0xe5e5d1d2, 0x1a1a2e2e, 0xededcbcc, 0x12123434, + 0xc9c9ebec, 0xd3d3d3d4, 0x2c2c2c2c, 0xc9c9dfe0, 0xd1d1d1d2, 0xd1d1d1d2, 0x2e2e2e2e, 0x00000000, + 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, + 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, + 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, + 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, + 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, + 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, + 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, + 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, + 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, + 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, + 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, + 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, + 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, + 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, + 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, + 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, + 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, + 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, + 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, + 0xf5f60000, 0x00000a0a, 0xfffff5f6, 0x00000000, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, + 0x04040404, 0xfbfbfbfc, 0x0a0a0a0a, 0xf5f5f5f6, 0x0a0a0000, 0xf5f60000, 0x00000a0a, 0xfffff5f6, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, + 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x0807f7f8, 0xf7f80808, 0xeff00808, + 0x100ff7f8, 0xe7e80000, 0x18180000, 0xf7f7e7e8, 0xf7f81818, 0x0807e7e8, 0x08081818, 0xeff01414, + 0x100febec, 0xffffe3e4, 0x00001c1c, 0xe7e7eff0, 0xe7e81010, 0x1817eff0, 0x18181010, 0xdfe00000, + 0x20200000, 0xefefe3e4, 0xeff01c1c, 0x100fe3e4, 0x10101c1c, 0xdfdff7f8, 0xdfe00808, 0xf7f7dfe0, + 0xf7f82020, 0x0807dfe0, 0x08082020, 0x201ff7f8, 0x20200808, 0x18181818, 0xe7e7e7e8, 0xe7e81818, + 0x1817e7e8, 0xdfdfebec, 0x20201414, 0xffffd7d8, 0x00002828, 0xefefd7d8, 0x10102828, 0xd3d40000, + 0xd3d40000, 0xffffd3d4, 0x00002c2c, 0x2c2c0000, 0x2c2c0000, 0xdfdfdfe0, 0x20202020, 0xd3d3eff0, + 0x2c2c1010, 0xd3d3e7e8, 0xe7e7d3d4, 0x18182c2c, 0x2c2c1818, 0xefefcfd0, 0x10103030, 0xdbdbdbdc, + 0xdbdbdbdc, 0x24242424, 0x24242424, 0xcbcbebec, 0x28282828, 0xd7d7d7d8, 0xcbcbdfe0, 0x00000000, + 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, + 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, + 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, + 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, + 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, + 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, + 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, + 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, + 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, + 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, + 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, + 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, + 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, + 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, + 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, + 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, + 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, + 0xf3f40000, 0x00000c0c, 0xfffff3f4, 0x00000000, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, + 0x04040404, 0xfbfbfbfc, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0000, 0xf3f40000, 0x00000c0c, 0xfffff3f4, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, + 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, + 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, + 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, + 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, + 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, + 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, + 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, + 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, + 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, + 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, + 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, + 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, + 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, + 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x06060606, 0xf9f9f9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x14141414, 0xebebebec, 0x20202020, 0xdfdfdfe0, 0x2e2e2e2e, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; static const uint32_t correctionhighorder[] = { - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, - 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, - 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, - 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, - 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, - 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, - 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, - 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, - 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, - 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, - 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, - 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, - 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, - 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, - 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, - 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, - 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, - 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, - 0x03030a0a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, - 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, - 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, - 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, - 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, - 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, - 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, - 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, - 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, - 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, - 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, - 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, - 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, - 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, - 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, - 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, - 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, - 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, - 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, - 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, - 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, - 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x09090909, 0xf6f6f6f7, - 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, - 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, - 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, - 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, - 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, - 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, - 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, - 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, - 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, - 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, - 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, - 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, - 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, - 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, - 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, - 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, - 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, - 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, - 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, - 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, - 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, - 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, - 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, - 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, - 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, - 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, - 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, - 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, - 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, - 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, - 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, - 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, - 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, - 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, - 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, - 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, - 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, - 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, - 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, - 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, - 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, - 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, - 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, - 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, - 0x04040808, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, - 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, - 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, - 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, - 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, - 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, - 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, - 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, - 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, - 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, - 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, - 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, - 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, - 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, - 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, - 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, - 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, - 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, - 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, - 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, - 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, - 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, - 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, - 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x09090909, 0xf6f6f6f7, - 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, - 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, - 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, - 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, - 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, - 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, - 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, - 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, - 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, - 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, - 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, - 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, - 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, - 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, - 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, - 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, - 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, - 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, - 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, - 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, - 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, - 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, - 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0x00000303, - 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, - 0x00000303, 0x00000303, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, - 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, - 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0xf8f90000, 0xf8f90000, - 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, - 0xf8f90000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, - 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, - 0x02020000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, - 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, - 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, - 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, - 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0x06060000, 0x06060000, 0x06060000, 0x06060000, - 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, - 0x06060000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, - 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0x00000606, 0x00000606, - 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, - 0x00000606, 0x00000606, 0x00000606, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, - 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, - 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, - 0x02020000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, - 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, - 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, - 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, - 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, - 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0xf5f5f5f6, - 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, - 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, - 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, - 0x0a0a0000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, - 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0x00000a0a, 0x00000a0a, - 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, - 0x00000a0a, 0x00000a0a, 0x00000a0a, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, - 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, - 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, - 0x04040000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, - 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0x00000404, 0x00000404, - 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, - 0x00000404, 0x00000404, 0x00000404, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, - 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, - 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, - 0xfbfbfbfc, 0xfbfbfbfc, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, - 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, - 0x0c0c0000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, - 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0x00000c0c, 0x00000c0c, - 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, - 0x00000c0c, 0x00000c0c, 0x00000c0c, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, - 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, - 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, - 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, - 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, - 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, - 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, - 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, - 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, - 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, - 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, - 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, - 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, - 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, - 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, - 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, - 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, + 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, + 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x0302feff, 0xfcfd0101, 0xfeff0303, 0x0100fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, + 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, + 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, + 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, + 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, + 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x0403feff, + 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, 0xfcfcfcfd, + 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, 0x03030303, + 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, 0x00000000, + 0x03030303, 0xfcfcfcfd, 0x0403feff, 0xfbfc0101, 0xfeff0404, 0x0100fbfc, 0x07070707, 0xf8f8f8f9, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, + 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, + 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, + 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, + 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, + 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, + 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, 0x03030a0a, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x0504feff, 0xfafb0101, 0xfeff0505, 0x0100fafb, 0x0a0a0303, 0xf5f5fcfd, + 0x03030a0a, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, + 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, + 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, + 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, + 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, + 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, + 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, + 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, + 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, + 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, + 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, + 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, + 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x05050505, + 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, + 0xf2f2fcfd, 0x00000000, 0x05050505, 0xfafafafb, 0x0706fdfe, 0xf8f90202, 0xfdfe0707, 0x0201f8f9, + 0x0b0b0b0b, 0xf4f4f4f5, 0x0d0d0303, 0xf2f2fcfd, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, + 0x04040f0f, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x0807fdfe, 0xf7f80202, 0xfdfe0808, 0x0201f7f8, + 0x0d0d0d0d, 0xf2f2f2f3, 0x0f0f0404, 0xf0f0fbfc, 0x04040f0f, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, + 0x05051212, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x0a09fcfd, 0xf5f60303, 0xfcfd0a0a, 0x0302f5f6, + 0x10101010, 0xefefeff0, 0x12120505, 0xededfafb, 0x05051212, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, + 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x0b0afcfd, 0xf4f50303, + 0xfcfd0b0b, 0x0302f4f5, 0x12121212, 0xedededee, 0x14140505, 0xebebfafb, 0x05051414, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x09090909, 0xf6f6f6f7, + 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, + 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, + 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, + 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, + 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, + 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, + 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, + 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, + 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, + 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, + 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, + 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, + 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, + 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, + 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, + 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, + 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, + 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, + 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, + 0xf9f9e8e9, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x0c0bfcfd, 0xf3f40303, 0xfcfd0c0c, 0x0302f3f4, + 0x14141414, 0xebebebec, 0x17170606, 0xe8e8f9fa, 0x06061717, 0xf9f9e8e9, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, + 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, + 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, + 0x00000000, 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, + 0x02020202, 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, + 0xfdfdfdfe, 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x02020202, 0xfdfdfdfe, + 0x02020000, 0xfdfe0000, 0x00000202, 0xfffffdfe, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, + 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, + 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, + 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, + 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, + 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, 0x03030000, + 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, 0xfcfcfcfd, + 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, 0x03030303, + 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, 0x00000000, + 0x03030303, 0xfcfcfcfd, 0x03030000, 0xfcfd0000, 0x00000303, 0xfffffcfd, 0x06060606, 0xf9f9f9fa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, + 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, + 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, + 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, + 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, + 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, + 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, 0x04040404, 0xfbfbfbfc, + 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, 0x04040808, 0x00000000, + 0x04040404, 0xfbfbfbfc, 0x04040000, 0xfbfc0000, 0x00000404, 0xfffffbfc, 0x08080404, 0xf7f7fbfc, + 0x04040808, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, + 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, + 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, + 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, + 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, + 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, + 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, + 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, + 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, + 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, + 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, + 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, + 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x05050505, + 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, + 0xf0f0fafb, 0x00000000, 0x05050505, 0xfafafafb, 0x05050000, 0xfafb0000, 0x00000505, 0xfffffafb, + 0x0a0a0a0a, 0xf5f5f5f6, 0x0f0f0505, 0xf0f0fafb, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, + 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, + 0x06060c0c, 0x00000000, 0x06060606, 0xf9f9f9fa, 0x06060000, 0xf9fa0000, 0x00000606, 0xfffff9fa, + 0x0c0c0c0c, 0xf3f3f3f4, 0x0c0c0606, 0xf3f3f9fa, 0x06060c0c, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, + 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, + 0x07071515, 0x00000000, 0x07070707, 0xf8f8f8f9, 0x07070000, 0xf8f90000, 0x00000707, 0xfffff8f9, + 0x0e0e0e0e, 0xf1f1f1f2, 0x15150707, 0xeaeaf8f9, 0x07071515, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, + 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, 0x08080808, 0xf7f7f7f8, 0x08080000, 0xf7f80000, + 0x00000808, 0xfffff7f8, 0x10101010, 0xefefeff0, 0x10100808, 0xefeff7f8, 0x08081010, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, 0x09090909, 0xf6f6f6f7, + 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, + 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, + 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, + 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, + 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, + 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, + 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, + 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, + 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, + 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, + 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, + 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, + 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, + 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, + 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, + 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, + 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, + 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, + 0xf6f70000, 0x00000909, 0xfffff6f7, 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, + 0xf6f6e4e5, 0x00000000, 0x09090909, 0xf6f6f6f7, 0x09090000, 0xf6f70000, 0x00000909, 0xfffff6f7, + 0x12121212, 0xedededee, 0x1b1b0909, 0xe4e4f6f7, 0x09091b1b, 0xf6f6e4e5, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0x03030000, + 0x03030000, 0x03030000, 0x03030000, 0x03030000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, + 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0xfcfd0000, 0x00000303, + 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, 0x00000303, + 0x00000303, 0x00000303, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, + 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0xfffffcfd, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, + 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0x07070000, 0xf8f90000, 0xf8f90000, + 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, 0xf8f90000, + 0xf8f90000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, + 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, + 0x02020000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, + 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, + 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, + 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, + 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0x06060000, 0x06060000, 0x06060000, 0x06060000, + 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, 0x06060000, + 0x06060000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, + 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0xf9fa0000, 0x00000606, 0x00000606, + 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, 0x00000606, + 0x00000606, 0x00000606, 0x00000606, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, + 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, 0xfffff9fa, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, + 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, 0x02020000, + 0x02020000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, + 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0xfdfe0000, 0x00000202, 0x00000202, + 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, 0x00000202, + 0x00000202, 0x00000202, 0x00000202, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, + 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, 0xfffffdfe, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, + 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0x0a0a0a0a, 0xf5f5f5f6, + 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, + 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0xf5f5f5f6, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, + 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, 0x0a0a0000, + 0x0a0a0000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, + 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0xf5f60000, 0x00000a0a, 0x00000a0a, + 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, 0x00000a0a, + 0x00000a0a, 0x00000a0a, 0x00000a0a, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, + 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, 0xfffff5f6, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, + 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, 0x04040000, + 0x04040000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, + 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0xfbfc0000, 0x00000404, 0x00000404, + 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, 0x00000404, + 0x00000404, 0x00000404, 0x00000404, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, + 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, 0xfffffbfc, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, + 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0x04040404, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, 0xfbfbfbfc, + 0xfbfbfbfc, 0xfbfbfbfc, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, + 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, 0x0c0c0000, + 0x0c0c0000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, + 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0xf3f40000, 0x00000c0c, 0x00000c0c, + 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, 0x00000c0c, + 0x00000c0c, 0x00000c0c, 0x00000c0c, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, + 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, 0xfffff3f4, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, + 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, + 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, + 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, + 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0xdeadbeef, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, 0x02020202, + 0x02020202, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, + 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0xfdfdfdfe, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, 0x06060606, + 0x06060606, 0x06060606, 0x06060606, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, 0xf9f9f9fa, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, + 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0x0c0c0c0c, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, 0xf3f3f3f4, + 0xf3f3f3f4, 0xf3f3f3f4, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, + 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0x14141414, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, + 0xebebebec, 0xebebebec, 0xebebebec, 0xebebebec, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, 0x20202020, + 0x20202020, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, + 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0xdfdfdfe0, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, + 0x2e2e2e2e, 0x2e2e2e2e, 0x2e2e2e2e, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, 0xd1d1d1d2, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 };
--- a/jfdctfst.c Wed Dec 21 17:50:40 2005 +0000 +++ b/jfdctfst.c Thu Dec 22 01:10:11 2005 +0000 @@ -83,10 +83,10 @@ */ #if CONST_BITS == 8 -#define FIX_0_382683433 ((int32_t) 98) /* FIX(0.382683433) */ -#define FIX_0_541196100 ((int32_t) 139) /* FIX(0.541196100) */ -#define FIX_0_707106781 ((int32_t) 181) /* FIX(0.707106781) */ -#define FIX_1_306562965 ((int32_t) 334) /* FIX(1.306562965) */ +#define FIX_0_382683433 ((int32_t) 98) /* FIX(0.382683433) */ +#define FIX_0_541196100 ((int32_t) 139) /* FIX(0.541196100) */ +#define FIX_0_707106781 ((int32_t) 181) /* FIX(0.707106781) */ +#define FIX_1_306562965 ((int32_t) 334) /* FIX(1.306562965) */ #else #define FIX_0_382683433 FIX(0.382683433) #define FIX_0_541196100 FIX(0.541196100) @@ -135,7 +135,7 @@ /* Even part */ - tmp10 = tmp0 + tmp3; /* phase 2 */ + tmp10 = tmp0 + tmp3; /* phase 2 */ tmp13 = tmp0 - tmp3; tmp11 = tmp1 + tmp2; tmp12 = tmp1 - tmp2; @@ -144,30 +144,30 @@ dataptr[4] = tmp10 - tmp11; z1 = MULTIPLY(tmp12 + tmp13, FIX_0_707106781); /* c4 */ - dataptr[2] = tmp13 + z1; /* phase 5 */ + dataptr[2] = tmp13 + z1; /* phase 5 */ dataptr[6] = tmp13 - z1; /* Odd part */ - tmp10 = tmp4 + tmp5; /* phase 2 */ + tmp10 = tmp4 + tmp5; /* phase 2 */ tmp11 = tmp5 + tmp6; tmp12 = tmp6 + tmp7; /* The rotator is modified from fig 4-8 to avoid extra negations. */ z5 = MULTIPLY(tmp10 - tmp12, FIX_0_382683433); /* c6 */ - z2 = MULTIPLY(tmp10, FIX_0_541196100) + z5; /* c2-c6 */ - z4 = MULTIPLY(tmp12, FIX_1_306562965) + z5; /* c2+c6 */ - z3 = MULTIPLY(tmp11, FIX_0_707106781); /* c4 */ + z2 = MULTIPLY(tmp10, FIX_0_541196100) + z5; /* c2-c6 */ + z4 = MULTIPLY(tmp12, FIX_1_306562965) + z5; /* c2+c6 */ + z3 = MULTIPLY(tmp11, FIX_0_707106781); /* c4 */ - z11 = tmp7 + z3; /* phase 5 */ + z11 = tmp7 + z3; /* phase 5 */ z13 = tmp7 - z3; - dataptr[5] = z13 + z2; /* phase 6 */ + dataptr[5] = z13 + z2; /* phase 6 */ dataptr[3] = z13 - z2; dataptr[1] = z11 + z4; dataptr[7] = z11 - z4; - dataptr += DCTSIZE; /* advance pointer to next row */ + dataptr += DCTSIZE; /* advance pointer to next row */ } } @@ -202,7 +202,7 @@ /* Even part */ - tmp10 = tmp0 + tmp3; /* phase 2 */ + tmp10 = tmp0 + tmp3; /* phase 2 */ tmp13 = tmp0 - tmp3; tmp11 = tmp1 + tmp2; tmp12 = tmp1 - tmp2; @@ -216,7 +216,7 @@ /* Odd part */ - tmp10 = tmp4 + tmp5; /* phase 2 */ + tmp10 = tmp4 + tmp5; /* phase 2 */ tmp11 = tmp5 + tmp6; tmp12 = tmp6 + tmp7; @@ -226,7 +226,7 @@ z4 = MULTIPLY(tmp12, FIX_1_306562965) + z5; /* c2+c6 */ z3 = MULTIPLY(tmp11, FIX_0_707106781); /* c4 */ - z11 = tmp7 + z3; /* phase 5 */ + z11 = tmp7 + z3; /* phase 5 */ z13 = tmp7 - z3; dataptr[DCTSIZE*5] = z13 + z2; /* phase 6 */ @@ -234,7 +234,7 @@ dataptr[DCTSIZE*1] = z11 + z4; dataptr[DCTSIZE*7] = z11 - z4; - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } } @@ -293,7 +293,7 @@ dataptr[DCTSIZE*3] = tmp13 + z1; dataptr[DCTSIZE*7] = tmp13 - z1; - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } }
--- a/jfdctint.c Wed Dec 21 17:50:40 2005 +0000 +++ b/jfdctint.c Thu Dec 22 01:10:11 2005 +0000 @@ -92,10 +92,10 @@ #if BITS_IN_JSAMPLE == 8 #define CONST_BITS 13 -#define PASS1_BITS 4 /* set this to 2 if 16x16 multiplies are faster */ +#define PASS1_BITS 4 /* set this to 2 if 16x16 multiplies are faster */ #else #define CONST_BITS 13 -#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ +#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ #endif /* Some C compilers fail to reduce "FIX(constant)" at compile time, thus @@ -106,18 +106,18 @@ */ #if CONST_BITS == 13 -#define FIX_0_298631336 ((int32_t) 2446) /* FIX(0.298631336) */ -#define FIX_0_390180644 ((int32_t) 3196) /* FIX(0.390180644) */ -#define FIX_0_541196100 ((int32_t) 4433) /* FIX(0.541196100) */ -#define FIX_0_765366865 ((int32_t) 6270) /* FIX(0.765366865) */ -#define FIX_0_899976223 ((int32_t) 7373) /* FIX(0.899976223) */ -#define FIX_1_175875602 ((int32_t) 9633) /* FIX(1.175875602) */ -#define FIX_1_501321110 ((int32_t) 12299) /* FIX(1.501321110) */ -#define FIX_1_847759065 ((int32_t) 15137) /* FIX(1.847759065) */ -#define FIX_1_961570560 ((int32_t) 16069) /* FIX(1.961570560) */ -#define FIX_2_053119869 ((int32_t) 16819) /* FIX(2.053119869) */ -#define FIX_2_562915447 ((int32_t) 20995) /* FIX(2.562915447) */ -#define FIX_3_072711026 ((int32_t) 25172) /* FIX(3.072711026) */ +#define FIX_0_298631336 ((int32_t) 2446) /* FIX(0.298631336) */ +#define FIX_0_390180644 ((int32_t) 3196) /* FIX(0.390180644) */ +#define FIX_0_541196100 ((int32_t) 4433) /* FIX(0.541196100) */ +#define FIX_0_765366865 ((int32_t) 6270) /* FIX(0.765366865) */ +#define FIX_0_899976223 ((int32_t) 7373) /* FIX(0.899976223) */ +#define FIX_1_175875602 ((int32_t) 9633) /* FIX(1.175875602) */ +#define FIX_1_501321110 ((int32_t) 12299) /* FIX(1.501321110) */ +#define FIX_1_847759065 ((int32_t) 15137) /* FIX(1.847759065) */ +#define FIX_1_961570560 ((int32_t) 16069) /* FIX(1.961570560) */ +#define FIX_2_053119869 ((int32_t) 16819) /* FIX(2.053119869) */ +#define FIX_2_562915447 ((int32_t) 20995) /* FIX(2.562915447) */ +#define FIX_3_072711026 ((int32_t) 25172) /* FIX(3.072711026) */ #else #define FIX_0_298631336 FIX(0.298631336) #define FIX_0_390180644 FIX(0.390180644) @@ -185,9 +185,9 @@ z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); dataptr[2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp13, FIX_0_765366865), - CONST_BITS-PASS1_BITS); + CONST_BITS-PASS1_BITS); dataptr[6] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, - FIX_1_847759065), - CONST_BITS-PASS1_BITS); + CONST_BITS-PASS1_BITS); /* Odd part per figure 8 --- note paper omits factor of sqrt(2). * cK represents cos(K*pi/16). @@ -217,7 +217,7 @@ dataptr[3] = (DCTELEM) DESCALE(tmp6 + z2 + z3, CONST_BITS-PASS1_BITS); dataptr[1] = (DCTELEM) DESCALE(tmp7 + z1 + z4, CONST_BITS-PASS1_BITS); - dataptr += DCTSIZE; /* advance pointer to next row */ + dataptr += DCTSIZE; /* advance pointer to next row */ } } @@ -267,9 +267,9 @@ z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp13, FIX_0_765366865), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, - FIX_1_847759065), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); /* Odd part per figure 8 --- note paper omits factor of sqrt(2). * cK represents cos(K*pi/16). @@ -295,15 +295,15 @@ z4 += z5; dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp4 + z1 + z3, - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp5 + z2 + z4, - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp6 + z2 + z3, - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp7 + z1 + z4, - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } } @@ -350,9 +350,9 @@ z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp13, FIX_0_765366865), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, - FIX_1_847759065), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); tmp10 = tmp4 + tmp7; tmp11 = tmp5 + tmp6; @@ -364,10 +364,10 @@ z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp13, FIX_0_765366865), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, - FIX_1_847759065), - CONST_BITS+PASS1_BITS); + CONST_BITS+PASS1_BITS); - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } }
--- a/jrevdct.c Wed Dec 21 17:50:40 2005 +0000 +++ b/jrevdct.c Thu Dec 22 01:10:11 2005 +0000 @@ -95,10 +95,10 @@ #ifdef EIGHT_BIT_SAMPLES #define PASS1_BITS 2 #else -#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ +#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ #endif -#define ONE ((int32_t) 1) +#define ONE ((int32_t) 1) #define CONST_SCALE (ONE << CONST_BITS) @@ -109,7 +109,7 @@ */ /* Actually FIX is no longer used, we precomputed them all */ -#define FIX(x) ((int32_t) ((x) * CONST_SCALE + 0.5)) +#define FIX(x) ((int32_t) ((x) * CONST_SCALE + 0.5)) /* Descale and correctly round an int32_t value that's scaled by N bits. * We assume RIGHT_SHIFT rounds towards minus infinity, so adding @@ -130,15 +130,15 @@ */ #ifdef EIGHT_BIT_SAMPLES -#ifdef SHORTxSHORT_32 /* may work if 'int' is 32 bits */ +#ifdef SHORTxSHORT_32 /* may work if 'int' is 32 bits */ #define MULTIPLY(var,const) (((int16_t) (var)) * ((int16_t) (const))) #endif -#ifdef SHORTxLCONST_32 /* known to work with Microsoft C 6.0 */ +#ifdef SHORTxLCONST_32 /* known to work with Microsoft C 6.0 */ #define MULTIPLY(var,const) (((int16_t) (var)) * ((int32_t) (const))) #endif #endif -#ifndef MULTIPLY /* default definition */ +#ifndef MULTIPLY /* default definition */ #define MULTIPLY(var,const) ((var) * (const)) #endif @@ -217,17 +217,17 @@ if ((d1 | d2 | d3 | d4 | d5 | d6 | d7) == 0) { /* AC terms all zero */ if (d0) { - /* Compute a 32 bit value to assign. */ - DCTELEM dcval = (DCTELEM) (d0 << PASS1_BITS); - register int v = (dcval & 0xffff) | ((dcval << 16) & 0xffff0000); + /* Compute a 32 bit value to assign. */ + DCTELEM dcval = (DCTELEM) (d0 << PASS1_BITS); + register int v = (dcval & 0xffff) | ((dcval << 16) & 0xffff0000); - idataptr[0] = v; - idataptr[1] = v; - idataptr[2] = v; - idataptr[3] = v; + idataptr[0] = v; + idataptr[1] = v; + idataptr[2] = v; + idataptr[3] = v; } - dataptr += DCTSIZE; /* advance pointer to next row */ + dataptr += DCTSIZE; /* advance pointer to next row */ continue; } @@ -235,50 +235,50 @@ /* The rotator is sqrt(2)*c(-6). */ { if (d6) { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ - z1 = MULTIPLY(d2 + d6, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); - tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ + z1 = MULTIPLY(d2 + d6, FIX_0_541196100); + tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); + tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ - tmp2 = MULTIPLY(-d6, FIX_1_306562965); - tmp3 = MULTIPLY(d6, FIX_0_541196100); + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ + tmp2 = MULTIPLY(-d6, FIX_1_306562965); + tmp3 = MULTIPLY(d6, FIX_0_541196100); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } } else { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ - tmp2 = MULTIPLY(d2, FIX_0_541196100); - tmp3 = MULTIPLY(d2, FIX_1_306562965); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ + tmp2 = MULTIPLY(d2, FIX_0_541196100); + tmp3 = MULTIPLY(d2, FIX_1_306562965); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ - tmp10 = tmp13 = (d0 + d4) << CONST_BITS; - tmp11 = tmp12 = (d0 - d4) << CONST_BITS; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ + tmp10 = tmp13 = (d0 + d4) << CONST_BITS; + tmp11 = tmp12 = (d0 - d4) << CONST_BITS; + } } /* Odd part per figure 8; the matrix is unitary and hence its @@ -286,259 +286,259 @@ */ if (d7) { - if (d5) { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 != 0, d7 != 0 */ - z1 = d7 + d1; - z2 = d5 + d3; - z3 = d7 + d3; - z4 = d5 + d1; - z5 = MULTIPLY(z3 + z4, FIX_1_175875602); + if (d5) { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 != 0, d7 != 0 */ + z1 = d7 + d1; + z2 = d5 + d3; + z3 = d7 + d3; + z4 = d5 + d1; + z5 = MULTIPLY(z3 + z4, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 != 0, d7 != 0 */ - z2 = d5 + d3; - z3 = d7 + d3; - z5 = MULTIPLY(z3 + d5, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 != 0, d7 != 0 */ + z2 = d5 + d3; + z3 = d7 + d3; + z5 = MULTIPLY(z3 + d5, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - z1 = MULTIPLY(-d7, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-d5, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + z1 = MULTIPLY(-d7, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-d5, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 = z1 + z4; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 != 0, d7 != 0 */ - z1 = d7 + d1; - z4 = d5 + d1; - z5 = MULTIPLY(d7 + z4, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 = z1 + z4; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 != 0, d7 != 0 */ + z1 = d7 + d1; + z4 = d5 + d1; + z5 = MULTIPLY(d7 + z4, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z3 = MULTIPLY(-d7, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z3 = MULTIPLY(-d7, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 = z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 == 0, d5 != 0, d7 != 0 */ - tmp0 = MULTIPLY(-d7, FIX_0_601344887); - z1 = MULTIPLY(-d7, FIX_0_899976223); - z3 = MULTIPLY(-d7, FIX_1_961570560); - tmp1 = MULTIPLY(-d5, FIX_0_509795579); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z4 = MULTIPLY(-d5, FIX_0_390180644); - z5 = MULTIPLY(d5 + d7, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 = z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 == 0, d5 != 0, d7 != 0 */ + tmp0 = MULTIPLY(-d7, FIX_0_601344887); + z1 = MULTIPLY(-d7, FIX_0_899976223); + z3 = MULTIPLY(-d7, FIX_1_961570560); + tmp1 = MULTIPLY(-d5, FIX_0_509795579); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z4 = MULTIPLY(-d5, FIX_0_390180644); + z5 = MULTIPLY(d5 + d7, FIX_1_175875602); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z3; - tmp1 += z4; - tmp2 = z2 + z3; - tmp3 = z1 + z4; - } - } - } else { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 == 0, d7 != 0 */ - z1 = d7 + d1; - z3 = d7 + d3; - z5 = MULTIPLY(z3 + d1, FIX_1_175875602); + tmp0 += z3; + tmp1 += z4; + tmp2 = z2 + z3; + tmp3 = z1 + z4; + } + } + } else { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 == 0, d7 != 0 */ + z1 = d7 + d1; + z3 = d7 + d3; + z5 = MULTIPLY(z3 + d1, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-d3, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-d1, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-d3, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-d1, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 = z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 == 0, d7 != 0 */ - z3 = d7 + d3; + tmp0 += z1 + z3; + tmp1 = z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 == 0, d7 != 0 */ + z3 = d7 + d3; - tmp0 = MULTIPLY(-d7, FIX_0_601344887); - z1 = MULTIPLY(-d7, FIX_0_899976223); - tmp2 = MULTIPLY(d3, FIX_0_509795579); - z2 = MULTIPLY(-d3, FIX_2_562915447); - z5 = MULTIPLY(z3, FIX_1_175875602); - z3 = MULTIPLY(-z3, FIX_0_785694958); + tmp0 = MULTIPLY(-d7, FIX_0_601344887); + z1 = MULTIPLY(-d7, FIX_0_899976223); + tmp2 = MULTIPLY(d3, FIX_0_509795579); + z2 = MULTIPLY(-d3, FIX_2_562915447); + z5 = MULTIPLY(z3, FIX_1_175875602); + z3 = MULTIPLY(-z3, FIX_0_785694958); - tmp0 += z3; - tmp1 = z2 + z5; - tmp2 += z3; - tmp3 = z1 + z5; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 == 0, d7 != 0 */ - z1 = d7 + d1; - z5 = MULTIPLY(z1, FIX_1_175875602); + tmp0 += z3; + tmp1 = z2 + z5; + tmp2 += z3; + tmp3 = z1 + z5; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 == 0, d7 != 0 */ + z1 = d7 + d1; + z5 = MULTIPLY(z1, FIX_1_175875602); - z1 = MULTIPLY(z1, FIX_0_275899380); - z3 = MULTIPLY(-d7, FIX_1_961570560); - tmp0 = MULTIPLY(-d7, FIX_1_662939225); - z4 = MULTIPLY(-d1, FIX_0_390180644); - tmp3 = MULTIPLY(d1, FIX_1_111140466); + z1 = MULTIPLY(z1, FIX_0_275899380); + z3 = MULTIPLY(-d7, FIX_1_961570560); + tmp0 = MULTIPLY(-d7, FIX_1_662939225); + z4 = MULTIPLY(-d1, FIX_0_390180644); + tmp3 = MULTIPLY(d1, FIX_1_111140466); - tmp0 += z1; - tmp1 = z4 + z5; - tmp2 = z3 + z5; - tmp3 += z1; - } else { - /* d1 == 0, d3 == 0, d5 == 0, d7 != 0 */ - tmp0 = MULTIPLY(-d7, FIX_1_387039845); - tmp1 = MULTIPLY(d7, FIX_1_175875602); - tmp2 = MULTIPLY(-d7, FIX_0_785694958); - tmp3 = MULTIPLY(d7, FIX_0_275899380); - } - } - } + tmp0 += z1; + tmp1 = z4 + z5; + tmp2 = z3 + z5; + tmp3 += z1; + } else { + /* d1 == 0, d3 == 0, d5 == 0, d7 != 0 */ + tmp0 = MULTIPLY(-d7, FIX_1_387039845); + tmp1 = MULTIPLY(d7, FIX_1_175875602); + tmp2 = MULTIPLY(-d7, FIX_0_785694958); + tmp3 = MULTIPLY(d7, FIX_0_275899380); + } + } + } } else { - if (d5) { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 != 0, d7 == 0 */ - z2 = d5 + d3; - z4 = d5 + d1; - z5 = MULTIPLY(d3 + z4, FIX_1_175875602); + if (d5) { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 != 0, d7 == 0 */ + z2 = d5 + d3; + z4 = d5 + d1; + z5 = MULTIPLY(d3 + z4, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-d1, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-d3, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-d1, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-d3, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 = z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 != 0, d7 == 0 */ - z2 = d5 + d3; + tmp0 = z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 != 0, d7 == 0 */ + z2 = d5 + d3; - z5 = MULTIPLY(z2, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_1_662939225); - z4 = MULTIPLY(-d5, FIX_0_390180644); - z2 = MULTIPLY(-z2, FIX_1_387039845); - tmp2 = MULTIPLY(d3, FIX_1_111140466); - z3 = MULTIPLY(-d3, FIX_1_961570560); + z5 = MULTIPLY(z2, FIX_1_175875602); + tmp1 = MULTIPLY(d5, FIX_1_662939225); + z4 = MULTIPLY(-d5, FIX_0_390180644); + z2 = MULTIPLY(-z2, FIX_1_387039845); + tmp2 = MULTIPLY(d3, FIX_1_111140466); + z3 = MULTIPLY(-d3, FIX_1_961570560); - tmp0 = z3 + z5; - tmp1 += z2; - tmp2 += z2; - tmp3 = z4 + z5; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 != 0, d7 == 0 */ - z4 = d5 + d1; + tmp0 = z3 + z5; + tmp1 += z2; + tmp2 += z2; + tmp3 = z4 + z5; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 != 0, d7 == 0 */ + z4 = d5 + d1; - z5 = MULTIPLY(z4, FIX_1_175875602); - z1 = MULTIPLY(-d1, FIX_0_899976223); - tmp3 = MULTIPLY(d1, FIX_0_601344887); - tmp1 = MULTIPLY(-d5, FIX_0_509795579); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z4 = MULTIPLY(z4, FIX_0_785694958); + z5 = MULTIPLY(z4, FIX_1_175875602); + z1 = MULTIPLY(-d1, FIX_0_899976223); + tmp3 = MULTIPLY(d1, FIX_0_601344887); + tmp1 = MULTIPLY(-d5, FIX_0_509795579); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z4 = MULTIPLY(z4, FIX_0_785694958); - tmp0 = z1 + z5; - tmp1 += z4; - tmp2 = z2 + z5; - tmp3 += z4; - } else { - /* d1 == 0, d3 == 0, d5 != 0, d7 == 0 */ - tmp0 = MULTIPLY(d5, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_0_275899380); - tmp2 = MULTIPLY(-d5, FIX_1_387039845); - tmp3 = MULTIPLY(d5, FIX_0_785694958); - } - } - } else { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 == 0, d7 == 0 */ - z5 = d1 + d3; - tmp3 = MULTIPLY(d1, FIX_0_211164243); - tmp2 = MULTIPLY(-d3, FIX_1_451774981); - z1 = MULTIPLY(d1, FIX_1_061594337); - z2 = MULTIPLY(-d3, FIX_2_172734803); - z4 = MULTIPLY(z5, FIX_0_785694958); - z5 = MULTIPLY(z5, FIX_1_175875602); + tmp0 = z1 + z5; + tmp1 += z4; + tmp2 = z2 + z5; + tmp3 += z4; + } else { + /* d1 == 0, d3 == 0, d5 != 0, d7 == 0 */ + tmp0 = MULTIPLY(d5, FIX_1_175875602); + tmp1 = MULTIPLY(d5, FIX_0_275899380); + tmp2 = MULTIPLY(-d5, FIX_1_387039845); + tmp3 = MULTIPLY(d5, FIX_0_785694958); + } + } + } else { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 == 0, d7 == 0 */ + z5 = d1 + d3; + tmp3 = MULTIPLY(d1, FIX_0_211164243); + tmp2 = MULTIPLY(-d3, FIX_1_451774981); + z1 = MULTIPLY(d1, FIX_1_061594337); + z2 = MULTIPLY(-d3, FIX_2_172734803); + z4 = MULTIPLY(z5, FIX_0_785694958); + z5 = MULTIPLY(z5, FIX_1_175875602); - tmp0 = z1 - z4; - tmp1 = z2 + z4; - tmp2 += z5; - tmp3 += z5; - } else { - /* d1 == 0, d3 != 0, d5 == 0, d7 == 0 */ - tmp0 = MULTIPLY(-d3, FIX_0_785694958); - tmp1 = MULTIPLY(-d3, FIX_1_387039845); - tmp2 = MULTIPLY(-d3, FIX_0_275899380); - tmp3 = MULTIPLY(d3, FIX_1_175875602); - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 == 0, d7 == 0 */ - tmp0 = MULTIPLY(d1, FIX_0_275899380); - tmp1 = MULTIPLY(d1, FIX_0_785694958); - tmp2 = MULTIPLY(d1, FIX_1_175875602); - tmp3 = MULTIPLY(d1, FIX_1_387039845); - } else { - /* d1 == 0, d3 == 0, d5 == 0, d7 == 0 */ - tmp0 = tmp1 = tmp2 = tmp3 = 0; - } - } - } + tmp0 = z1 - z4; + tmp1 = z2 + z4; + tmp2 += z5; + tmp3 += z5; + } else { + /* d1 == 0, d3 != 0, d5 == 0, d7 == 0 */ + tmp0 = MULTIPLY(-d3, FIX_0_785694958); + tmp1 = MULTIPLY(-d3, FIX_1_387039845); + tmp2 = MULTIPLY(-d3, FIX_0_275899380); + tmp3 = MULTIPLY(d3, FIX_1_175875602); + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 == 0, d7 == 0 */ + tmp0 = MULTIPLY(d1, FIX_0_275899380); + tmp1 = MULTIPLY(d1, FIX_0_785694958); + tmp2 = MULTIPLY(d1, FIX_1_175875602); + tmp3 = MULTIPLY(d1, FIX_1_387039845); + } else { + /* d1 == 0, d3 == 0, d5 == 0, d7 == 0 */ + tmp0 = tmp1 = tmp2 = tmp3 = 0; + } + } + } } } /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ @@ -552,7 +552,7 @@ dataptr[3] = (DCTELEM) DESCALE(tmp13 + tmp0, CONST_BITS-PASS1_BITS); dataptr[4] = (DCTELEM) DESCALE(tmp13 - tmp0, CONST_BITS-PASS1_BITS); - dataptr += DCTSIZE; /* advance pointer to next row */ + dataptr += DCTSIZE; /* advance pointer to next row */ } /* Pass 2: process columns. */ @@ -581,334 +581,334 @@ /* Even part: reverse the even part of the forward DCT. */ /* The rotator is sqrt(2)*c(-6). */ if (d6) { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ - z1 = MULTIPLY(d2 + d6, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); - tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ + z1 = MULTIPLY(d2 + d6, FIX_0_541196100); + tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); + tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ - tmp2 = MULTIPLY(-d6, FIX_1_306562965); - tmp3 = MULTIPLY(d6, FIX_0_541196100); + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ + tmp2 = MULTIPLY(-d6, FIX_1_306562965); + tmp3 = MULTIPLY(d6, FIX_0_541196100); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } } else { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ - tmp2 = MULTIPLY(d2, FIX_0_541196100); - tmp3 = MULTIPLY(d2, FIX_1_306562965); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ + tmp2 = MULTIPLY(d2, FIX_0_541196100); + tmp3 = MULTIPLY(d2, FIX_1_306562965); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ - tmp10 = tmp13 = (d0 + d4) << CONST_BITS; - tmp11 = tmp12 = (d0 - d4) << CONST_BITS; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ + tmp10 = tmp13 = (d0 + d4) << CONST_BITS; + tmp11 = tmp12 = (d0 - d4) << CONST_BITS; + } } /* Odd part per figure 8; the matrix is unitary and hence its * transpose is its inverse. i0..i3 are y7,y5,y3,y1 respectively. */ if (d7) { - if (d5) { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 != 0, d7 != 0 */ - z1 = d7 + d1; - z2 = d5 + d3; - z3 = d7 + d3; - z4 = d5 + d1; - z5 = MULTIPLY(z3 + z4, FIX_1_175875602); + if (d5) { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 != 0, d7 != 0 */ + z1 = d7 + d1; + z2 = d5 + d3; + z3 = d7 + d3; + z4 = d5 + d1; + z5 = MULTIPLY(z3 + z4, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 != 0, d7 != 0 */ - z1 = d7; - z2 = d5 + d3; - z3 = d7 + d3; - z5 = MULTIPLY(z3 + d5, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 != 0, d7 != 0 */ + z1 = d7; + z2 = d5 + d3; + z3 = d7 + d3; + z5 = MULTIPLY(z3 + d5, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - z1 = MULTIPLY(-d7, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-d5, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + z1 = MULTIPLY(-d7, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-d5, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 = z1 + z4; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 != 0, d7 != 0 */ - z1 = d7 + d1; - z2 = d5; - z3 = d7; - z4 = d5 + d1; - z5 = MULTIPLY(z3 + z4, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 = z1 + z4; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 != 0, d7 != 0 */ + z1 = d7 + d1; + z2 = d5; + z3 = d7; + z4 = d5 + d1; + z5 = MULTIPLY(z3 + z4, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z3 = MULTIPLY(-d7, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z3 = MULTIPLY(-d7, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 += z2 + z4; - tmp2 = z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 == 0, d5 != 0, d7 != 0 */ - tmp0 = MULTIPLY(-d7, FIX_0_601344887); - z1 = MULTIPLY(-d7, FIX_0_899976223); - z3 = MULTIPLY(-d7, FIX_1_961570560); - tmp1 = MULTIPLY(-d5, FIX_0_509795579); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z4 = MULTIPLY(-d5, FIX_0_390180644); - z5 = MULTIPLY(d5 + d7, FIX_1_175875602); + tmp0 += z1 + z3; + tmp1 += z2 + z4; + tmp2 = z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 == 0, d5 != 0, d7 != 0 */ + tmp0 = MULTIPLY(-d7, FIX_0_601344887); + z1 = MULTIPLY(-d7, FIX_0_899976223); + z3 = MULTIPLY(-d7, FIX_1_961570560); + tmp1 = MULTIPLY(-d5, FIX_0_509795579); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z4 = MULTIPLY(-d5, FIX_0_390180644); + z5 = MULTIPLY(d5 + d7, FIX_1_175875602); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z3; - tmp1 += z4; - tmp2 = z2 + z3; - tmp3 = z1 + z4; - } - } - } else { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 == 0, d7 != 0 */ - z1 = d7 + d1; - z3 = d7 + d3; - z5 = MULTIPLY(z3 + d1, FIX_1_175875602); + tmp0 += z3; + tmp1 += z4; + tmp2 = z2 + z3; + tmp3 = z1 + z4; + } + } + } else { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 == 0, d7 != 0 */ + z1 = d7 + d1; + z3 = d7 + d3; + z5 = MULTIPLY(z3 + d1, FIX_1_175875602); - tmp0 = MULTIPLY(d7, FIX_0_298631336); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-z1, FIX_0_899976223); - z2 = MULTIPLY(-d3, FIX_2_562915447); - z3 = MULTIPLY(-z3, FIX_1_961570560); - z4 = MULTIPLY(-d1, FIX_0_390180644); + tmp0 = MULTIPLY(d7, FIX_0_298631336); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-z1, FIX_0_899976223); + z2 = MULTIPLY(-d3, FIX_2_562915447); + z3 = MULTIPLY(-z3, FIX_1_961570560); + z4 = MULTIPLY(-d1, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 += z1 + z3; - tmp1 = z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 == 0, d7 != 0 */ - z3 = d7 + d3; + tmp0 += z1 + z3; + tmp1 = z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 == 0, d7 != 0 */ + z3 = d7 + d3; - tmp0 = MULTIPLY(-d7, FIX_0_601344887); - z1 = MULTIPLY(-d7, FIX_0_899976223); - tmp2 = MULTIPLY(d3, FIX_0_509795579); - z2 = MULTIPLY(-d3, FIX_2_562915447); - z5 = MULTIPLY(z3, FIX_1_175875602); - z3 = MULTIPLY(-z3, FIX_0_785694958); + tmp0 = MULTIPLY(-d7, FIX_0_601344887); + z1 = MULTIPLY(-d7, FIX_0_899976223); + tmp2 = MULTIPLY(d3, FIX_0_509795579); + z2 = MULTIPLY(-d3, FIX_2_562915447); + z5 = MULTIPLY(z3, FIX_1_175875602); + z3 = MULTIPLY(-z3, FIX_0_785694958); - tmp0 += z3; - tmp1 = z2 + z5; - tmp2 += z3; - tmp3 = z1 + z5; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 == 0, d7 != 0 */ - z1 = d7 + d1; - z5 = MULTIPLY(z1, FIX_1_175875602); + tmp0 += z3; + tmp1 = z2 + z5; + tmp2 += z3; + tmp3 = z1 + z5; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 == 0, d7 != 0 */ + z1 = d7 + d1; + z5 = MULTIPLY(z1, FIX_1_175875602); - z1 = MULTIPLY(z1, FIX_0_275899380); - z3 = MULTIPLY(-d7, FIX_1_961570560); - tmp0 = MULTIPLY(-d7, FIX_1_662939225); - z4 = MULTIPLY(-d1, FIX_0_390180644); - tmp3 = MULTIPLY(d1, FIX_1_111140466); + z1 = MULTIPLY(z1, FIX_0_275899380); + z3 = MULTIPLY(-d7, FIX_1_961570560); + tmp0 = MULTIPLY(-d7, FIX_1_662939225); + z4 = MULTIPLY(-d1, FIX_0_390180644); + tmp3 = MULTIPLY(d1, FIX_1_111140466); - tmp0 += z1; - tmp1 = z4 + z5; - tmp2 = z3 + z5; - tmp3 += z1; - } else { - /* d1 == 0, d3 == 0, d5 == 0, d7 != 0 */ - tmp0 = MULTIPLY(-d7, FIX_1_387039845); - tmp1 = MULTIPLY(d7, FIX_1_175875602); - tmp2 = MULTIPLY(-d7, FIX_0_785694958); - tmp3 = MULTIPLY(d7, FIX_0_275899380); - } - } - } + tmp0 += z1; + tmp1 = z4 + z5; + tmp2 = z3 + z5; + tmp3 += z1; + } else { + /* d1 == 0, d3 == 0, d5 == 0, d7 != 0 */ + tmp0 = MULTIPLY(-d7, FIX_1_387039845); + tmp1 = MULTIPLY(d7, FIX_1_175875602); + tmp2 = MULTIPLY(-d7, FIX_0_785694958); + tmp3 = MULTIPLY(d7, FIX_0_275899380); + } + } + } } else { - if (d5) { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 != 0, d7 == 0 */ - z2 = d5 + d3; - z4 = d5 + d1; - z5 = MULTIPLY(d3 + z4, FIX_1_175875602); + if (d5) { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 != 0, d7 == 0 */ + z2 = d5 + d3; + z4 = d5 + d1; + z5 = MULTIPLY(d3 + z4, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_2_053119869); - tmp2 = MULTIPLY(d3, FIX_3_072711026); - tmp3 = MULTIPLY(d1, FIX_1_501321110); - z1 = MULTIPLY(-d1, FIX_0_899976223); - z2 = MULTIPLY(-z2, FIX_2_562915447); - z3 = MULTIPLY(-d3, FIX_1_961570560); - z4 = MULTIPLY(-z4, FIX_0_390180644); + tmp1 = MULTIPLY(d5, FIX_2_053119869); + tmp2 = MULTIPLY(d3, FIX_3_072711026); + tmp3 = MULTIPLY(d1, FIX_1_501321110); + z1 = MULTIPLY(-d1, FIX_0_899976223); + z2 = MULTIPLY(-z2, FIX_2_562915447); + z3 = MULTIPLY(-d3, FIX_1_961570560); + z4 = MULTIPLY(-z4, FIX_0_390180644); - z3 += z5; - z4 += z5; + z3 += z5; + z4 += z5; - tmp0 = z1 + z3; - tmp1 += z2 + z4; - tmp2 += z2 + z3; - tmp3 += z1 + z4; - } else { - /* d1 == 0, d3 != 0, d5 != 0, d7 == 0 */ - z2 = d5 + d3; + tmp0 = z1 + z3; + tmp1 += z2 + z4; + tmp2 += z2 + z3; + tmp3 += z1 + z4; + } else { + /* d1 == 0, d3 != 0, d5 != 0, d7 == 0 */ + z2 = d5 + d3; - z5 = MULTIPLY(z2, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_1_662939225); - z4 = MULTIPLY(-d5, FIX_0_390180644); - z2 = MULTIPLY(-z2, FIX_1_387039845); - tmp2 = MULTIPLY(d3, FIX_1_111140466); - z3 = MULTIPLY(-d3, FIX_1_961570560); + z5 = MULTIPLY(z2, FIX_1_175875602); + tmp1 = MULTIPLY(d5, FIX_1_662939225); + z4 = MULTIPLY(-d5, FIX_0_390180644); + z2 = MULTIPLY(-z2, FIX_1_387039845); + tmp2 = MULTIPLY(d3, FIX_1_111140466); + z3 = MULTIPLY(-d3, FIX_1_961570560); - tmp0 = z3 + z5; - tmp1 += z2; - tmp2 += z2; - tmp3 = z4 + z5; - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 != 0, d7 == 0 */ - z4 = d5 + d1; + tmp0 = z3 + z5; + tmp1 += z2; + tmp2 += z2; + tmp3 = z4 + z5; + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 != 0, d7 == 0 */ + z4 = d5 + d1; - z5 = MULTIPLY(z4, FIX_1_175875602); - z1 = MULTIPLY(-d1, FIX_0_899976223); - tmp3 = MULTIPLY(d1, FIX_0_601344887); - tmp1 = MULTIPLY(-d5, FIX_0_509795579); - z2 = MULTIPLY(-d5, FIX_2_562915447); - z4 = MULTIPLY(z4, FIX_0_785694958); + z5 = MULTIPLY(z4, FIX_1_175875602); + z1 = MULTIPLY(-d1, FIX_0_899976223); + tmp3 = MULTIPLY(d1, FIX_0_601344887); + tmp1 = MULTIPLY(-d5, FIX_0_509795579); + z2 = MULTIPLY(-d5, FIX_2_562915447); + z4 = MULTIPLY(z4, FIX_0_785694958); - tmp0 = z1 + z5; - tmp1 += z4; - tmp2 = z2 + z5; - tmp3 += z4; - } else { - /* d1 == 0, d3 == 0, d5 != 0, d7 == 0 */ - tmp0 = MULTIPLY(d5, FIX_1_175875602); - tmp1 = MULTIPLY(d5, FIX_0_275899380); - tmp2 = MULTIPLY(-d5, FIX_1_387039845); - tmp3 = MULTIPLY(d5, FIX_0_785694958); - } - } - } else { - if (d3) { - if (d1) { - /* d1 != 0, d3 != 0, d5 == 0, d7 == 0 */ - z5 = d1 + d3; - tmp3 = MULTIPLY(d1, FIX_0_211164243); - tmp2 = MULTIPLY(-d3, FIX_1_451774981); - z1 = MULTIPLY(d1, FIX_1_061594337); - z2 = MULTIPLY(-d3, FIX_2_172734803); - z4 = MULTIPLY(z5, FIX_0_785694958); - z5 = MULTIPLY(z5, FIX_1_175875602); + tmp0 = z1 + z5; + tmp1 += z4; + tmp2 = z2 + z5; + tmp3 += z4; + } else { + /* d1 == 0, d3 == 0, d5 != 0, d7 == 0 */ + tmp0 = MULTIPLY(d5, FIX_1_175875602); + tmp1 = MULTIPLY(d5, FIX_0_275899380); + tmp2 = MULTIPLY(-d5, FIX_1_387039845); + tmp3 = MULTIPLY(d5, FIX_0_785694958); + } + } + } else { + if (d3) { + if (d1) { + /* d1 != 0, d3 != 0, d5 == 0, d7 == 0 */ + z5 = d1 + d3; + tmp3 = MULTIPLY(d1, FIX_0_211164243); + tmp2 = MULTIPLY(-d3, FIX_1_451774981); + z1 = MULTIPLY(d1, FIX_1_061594337); + z2 = MULTIPLY(-d3, FIX_2_172734803); + z4 = MULTIPLY(z5, FIX_0_785694958); + z5 = MULTIPLY(z5, FIX_1_175875602); - tmp0 = z1 - z4; - tmp1 = z2 + z4; - tmp2 += z5; - tmp3 += z5; - } else { - /* d1 == 0, d3 != 0, d5 == 0, d7 == 0 */ - tmp0 = MULTIPLY(-d3, FIX_0_785694958); - tmp1 = MULTIPLY(-d3, FIX_1_387039845); - tmp2 = MULTIPLY(-d3, FIX_0_275899380); - tmp3 = MULTIPLY(d3, FIX_1_175875602); - } - } else { - if (d1) { - /* d1 != 0, d3 == 0, d5 == 0, d7 == 0 */ - tmp0 = MULTIPLY(d1, FIX_0_275899380); - tmp1 = MULTIPLY(d1, FIX_0_785694958); - tmp2 = MULTIPLY(d1, FIX_1_175875602); - tmp3 = MULTIPLY(d1, FIX_1_387039845); - } else { - /* d1 == 0, d3 == 0, d5 == 0, d7 == 0 */ - tmp0 = tmp1 = tmp2 = tmp3 = 0; - } - } - } + tmp0 = z1 - z4; + tmp1 = z2 + z4; + tmp2 += z5; + tmp3 += z5; + } else { + /* d1 == 0, d3 != 0, d5 == 0, d7 == 0 */ + tmp0 = MULTIPLY(-d3, FIX_0_785694958); + tmp1 = MULTIPLY(-d3, FIX_1_387039845); + tmp2 = MULTIPLY(-d3, FIX_0_275899380); + tmp3 = MULTIPLY(d3, FIX_1_175875602); + } + } else { + if (d1) { + /* d1 != 0, d3 == 0, d5 == 0, d7 == 0 */ + tmp0 = MULTIPLY(d1, FIX_0_275899380); + tmp1 = MULTIPLY(d1, FIX_0_785694958); + tmp2 = MULTIPLY(d1, FIX_1_175875602); + tmp3 = MULTIPLY(d1, FIX_1_387039845); + } else { + /* d1 == 0, d3 == 0, d5 == 0, d7 == 0 */ + tmp0 = tmp1 = tmp2 = tmp3 = 0; + } + } + } } /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ dataptr[DCTSIZE*0] = (DCTELEM) DESCALE(tmp10 + tmp3, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp10 - tmp3, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp11 + tmp2, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(tmp11 - tmp2, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(tmp12 + tmp1, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12 - tmp1, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp13 + tmp0, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(tmp13 - tmp0, - CONST_BITS+PASS1_BITS+3); + CONST_BITS+PASS1_BITS+3); - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } } @@ -953,65 +953,65 @@ if ((d2 | d4 | d6) == 0) { /* AC terms all zero */ if (d0) { - /* Compute a 32 bit value to assign. */ - DCTELEM dcval = (DCTELEM) (d0 << PASS1_BITS); - register int v = (dcval & 0xffff) | ((dcval << 16) & 0xffff0000); + /* Compute a 32 bit value to assign. */ + DCTELEM dcval = (DCTELEM) (d0 << PASS1_BITS); + register int v = (dcval & 0xffff) | ((dcval << 16) & 0xffff0000); - idataptr[0] = v; - idataptr[1] = v; + idataptr[0] = v; + idataptr[1] = v; } - dataptr += DCTSTRIDE; /* advance pointer to next row */ + dataptr += DCTSTRIDE; /* advance pointer to next row */ continue; } /* Even part: reverse the even part of the forward DCT. */ /* The rotator is sqrt(2)*c(-6). */ if (d6) { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ - z1 = MULTIPLY(d2 + d6, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); - tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ + z1 = MULTIPLY(d2 + d6, FIX_0_541196100); + tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); + tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ - tmp2 = MULTIPLY(-d6, FIX_1_306562965); - tmp3 = MULTIPLY(d6, FIX_0_541196100); + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ + tmp2 = MULTIPLY(-d6, FIX_1_306562965); + tmp3 = MULTIPLY(d6, FIX_0_541196100); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } } else { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ - tmp2 = MULTIPLY(d2, FIX_0_541196100); - tmp3 = MULTIPLY(d2, FIX_1_306562965); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ + tmp2 = MULTIPLY(d2, FIX_0_541196100); + tmp3 = MULTIPLY(d2, FIX_1_306562965); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ - tmp10 = tmp13 = (d0 + d4) << CONST_BITS; - tmp11 = tmp12 = (d0 - d4) << CONST_BITS; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ + tmp10 = tmp13 = (d0 + d4) << CONST_BITS; + tmp11 = tmp12 = (d0 - d4) << CONST_BITS; + } } /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ @@ -1021,7 +1021,7 @@ dataptr[2] = (DCTELEM) DESCALE(tmp12, CONST_BITS-PASS1_BITS); dataptr[3] = (DCTELEM) DESCALE(tmp13, CONST_BITS-PASS1_BITS); - dataptr += DCTSTRIDE; /* advance pointer to next row */ + dataptr += DCTSTRIDE; /* advance pointer to next row */ } /* Pass 2: process columns. */ @@ -1046,50 +1046,50 @@ /* Even part: reverse the even part of the forward DCT. */ /* The rotator is sqrt(2)*c(-6). */ if (d6) { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ - z1 = MULTIPLY(d2 + d6, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); - tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 != 0 */ + z1 = MULTIPLY(d2 + d6, FIX_0_541196100); + tmp2 = z1 + MULTIPLY(-d6, FIX_1_847759065); + tmp3 = z1 + MULTIPLY(d2, FIX_0_765366865); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ - tmp2 = MULTIPLY(-d6, FIX_1_306562965); - tmp3 = MULTIPLY(d6, FIX_0_541196100); + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 != 0 */ + tmp2 = MULTIPLY(-d6, FIX_1_306562965); + tmp3 = MULTIPLY(d6, FIX_0_541196100); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } } else { - if (d2) { - /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ - tmp2 = MULTIPLY(d2, FIX_0_541196100); - tmp3 = MULTIPLY(d2, FIX_1_306562965); + if (d2) { + /* d0 != 0, d2 != 0, d4 != 0, d6 == 0 */ + tmp2 = MULTIPLY(d2, FIX_0_541196100); + tmp3 = MULTIPLY(d2, FIX_1_306562965); - tmp0 = (d0 + d4) << CONST_BITS; - tmp1 = (d0 - d4) << CONST_BITS; + tmp0 = (d0 + d4) << CONST_BITS; + tmp1 = (d0 - d4) << CONST_BITS; - tmp10 = tmp0 + tmp3; - tmp13 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp1 - tmp2; - } else { - /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ - tmp10 = tmp13 = (d0 + d4) << CONST_BITS; - tmp11 = tmp12 = (d0 - d4) << CONST_BITS; - } + tmp10 = tmp0 + tmp3; + tmp13 = tmp0 - tmp3; + tmp11 = tmp1 + tmp2; + tmp12 = tmp1 - tmp2; + } else { + /* d0 != 0, d2 == 0, d4 != 0, d6 == 0 */ + tmp10 = tmp13 = (d0 + d4) << CONST_BITS; + tmp11 = tmp12 = (d0 - d4) << CONST_BITS; + } } /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ @@ -1099,7 +1099,7 @@ dataptr[DCTSTRIDE*2] = tmp12 >> (CONST_BITS+PASS1_BITS+3); dataptr[DCTSTRIDE*3] = tmp13 >> (CONST_BITS+PASS1_BITS+3); - dataptr++; /* advance pointer to next column */ + dataptr++; /* advance pointer to next column */ } }
--- a/lcl.c Wed Dec 21 17:50:40 2005 +0000 +++ b/lcl.c Thu Dec 22 01:10:11 2005 +0000 @@ -81,8 +81,8 @@ */ typedef struct LclContext { - AVCodecContext *avctx; - AVFrame pic; + AVCodecContext *avctx; + AVFrame pic; PutBitContext pb; // Image type @@ -198,8 +198,8 @@ */ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8_t *buf, int buf_size) { - LclContext * const c = (LclContext *)avctx->priv_data; - unsigned char *encoded = (unsigned char *)buf; + LclContext * const c = (LclContext *)avctx->priv_data; + unsigned char *encoded = (unsigned char *)buf; unsigned int pixel_ptr; int row, col; unsigned char *outptr; @@ -214,15 +214,15 @@ #endif unsigned int len = buf_size; - if(c->pic.data[0]) - avctx->release_buffer(avctx, &c->pic); + if(c->pic.data[0]) + avctx->release_buffer(avctx, &c->pic); - c->pic.reference = 0; - c->pic.buffer_hints = FF_BUFFER_HINTS_VALID; - if(avctx->get_buffer(avctx, &c->pic) < 0){ - av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); - return -1; - } + c->pic.reference = 0; + c->pic.buffer_hints = FF_BUFFER_HINTS_VALID; + if(avctx->get_buffer(avctx, &c->pic) < 0){ + av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); + return -1; + } outptr = c->pic.data[0]; // Output image pointer @@ -358,7 +358,7 @@ pixel_ptr = row * width * 3; yq = encoded[pixel_ptr++]; uqvq = encoded[pixel_ptr++]; - uqvq+=(encoded[pixel_ptr++] << 8); + uqvq+=(encoded[pixel_ptr++] << 8); for (col = 1; col < width; col++) { encoded[pixel_ptr] = yq -= encoded[pixel_ptr]; uqvq -= (encoded[pixel_ptr+1] | (encoded[pixel_ptr+2]<<8)); @@ -588,8 +588,8 @@ c->zstream.avail_in = avctx->width*3; zret = deflate(&(c->zstream), Z_NO_FLUSH); if (zret != Z_OK) { - av_log(avctx, AV_LOG_ERROR, "Deflate error: %d\n", zret); - return -1; + av_log(avctx, AV_LOG_ERROR, "Deflate error: %d\n", zret); + return -1; } } zret = deflate(&(c->zstream), Z_FINISH); @@ -714,7 +714,7 @@ break; default: if ((c->compression < Z_NO_COMPRESSION) || (c->compression > Z_BEST_COMPRESSION)) { - av_log(avctx, AV_LOG_ERROR, "Unsupported compression level for ZLIB: (%d).\n", c->compression); + av_log(avctx, AV_LOG_ERROR, "Unsupported compression level for ZLIB: (%d).\n", c->compression); return 1; } av_log(avctx, AV_LOG_INFO, "Compression level for ZLIB: (%d).\n", c->compression); @@ -851,15 +851,15 @@ */ static int decode_end(AVCodecContext *avctx) { - LclContext * const c = (LclContext *)avctx->priv_data; + LclContext * const c = (LclContext *)avctx->priv_data; - if (c->pic.data[0]) - avctx->release_buffer(avctx, &c->pic); + if (c->pic.data[0]) + avctx->release_buffer(avctx, &c->pic); #ifdef CONFIG_ZLIB inflateEnd(&(c->zstream)); #endif - return 0; + return 0; } @@ -883,28 +883,28 @@ } AVCodec mszh_decoder = { - "mszh", - CODEC_TYPE_VIDEO, - CODEC_ID_MSZH, - sizeof(LclContext), - decode_init, - NULL, - decode_end, - decode_frame, - CODEC_CAP_DR1, + "mszh", + CODEC_TYPE_VIDEO, + CODEC_ID_MSZH, + sizeof(LclContext), + decode_init, + NULL, + decode_end, + decode_frame, + CODEC_CAP_DR1, }; AVCodec zlib_decoder = { - "zlib", - CODEC_TYPE_VIDEO, - CODEC_ID_ZLIB, - sizeof(LclContext), - decode_init, - NULL, - decode_end, - decode_frame, - CODEC_CAP_DR1, + "zlib", + CODEC_TYPE_VIDEO, + CODEC_ID_ZLIB, + sizeof(LclContext), + decode_init, + NULL, + decode_end, + decode_frame, + CODEC_CAP_DR1, }; #ifdef CONFIG_ENCODERS
--- a/libpostproc/postprocess.c Wed Dec 21 17:50:40 2005 +0000 +++ b/libpostproc/postprocess.c Thu Dec 22 01:10:11 2005 +0000 @@ -24,26 +24,26 @@ */ /* - C MMX MMX2 3DNow AltiVec -isVertDC Ec Ec Ec -isVertMinMaxOk Ec Ec Ec -doVertLowPass E e e Ec -doVertDefFilter Ec Ec e e Ec -isHorizDC Ec Ec Ec -isHorizMinMaxOk a E Ec -doHorizLowPass E e e Ec -doHorizDefFilter Ec Ec e e Ec -do_a_deblock Ec E Ec E -deRing E e e* Ecp -Vertical RKAlgo1 E a a -Horizontal RKAlgo1 a a -Vertical X1# a E E -Horizontal X1# a E E -LinIpolDeinterlace e E E* -CubicIpolDeinterlace a e e* -LinBlendDeinterlace e E E* -MedianDeinterlace# E Ec Ec -TempDeNoiser# E e e Ec + C MMX MMX2 3DNow AltiVec +isVertDC Ec Ec Ec +isVertMinMaxOk Ec Ec Ec +doVertLowPass E e e Ec +doVertDefFilter Ec Ec e e Ec +isHorizDC Ec Ec Ec +isHorizMinMaxOk a E Ec +doHorizLowPass E e e Ec +doHorizDefFilter Ec Ec e e Ec +do_a_deblock Ec E Ec E +deRing E e e* Ecp +Vertical RKAlgo1 E a a +Horizontal RKAlgo1 a a +Vertical X1# a E E +Horizontal X1# a E E +LinIpolDeinterlace e E E* +CubicIpolDeinterlace a e e* +LinBlendDeinterlace e E E* +MedianDeinterlace# E Ec Ec +TempDeNoiser# E e e Ec * i dont have a 3dnow CPU -> its untested, but noone said it doesnt work so it seems to work # more or less selfinvented filters so the exactness isnt too meaningfull @@ -61,7 +61,7 @@ move YScale thing to the end instead of fixing QP write a faster and higher quality deblocking filter :) make the mainloop more flexible (variable number of blocks at once - (the if/else stuff per block is slowing things down) + (the if/else stuff per block is slowing things down) compare the quality & speed of all filters split this huge file optimize c versions @@ -120,14 +120,14 @@ #endif #if defined(ARCH_X86) || defined(ARCH_X86_64) -static uint64_t __attribute__((aligned(8))) attribute_used w05= 0x0005000500050005LL; -static uint64_t __attribute__((aligned(8))) attribute_used w04= 0x0004000400040004LL; -static uint64_t __attribute__((aligned(8))) attribute_used w20= 0x0020002000200020LL; -static uint64_t __attribute__((aligned(8))) attribute_used b00= 0x0000000000000000LL; -static uint64_t __attribute__((aligned(8))) attribute_used b01= 0x0101010101010101LL; -static uint64_t __attribute__((aligned(8))) attribute_used b02= 0x0202020202020202LL; -static uint64_t __attribute__((aligned(8))) attribute_used b08= 0x0808080808080808LL; -static uint64_t __attribute__((aligned(8))) attribute_used b80= 0x8080808080808080LL; +static uint64_t __attribute__((aligned(8))) attribute_used w05= 0x0005000500050005LL; +static uint64_t __attribute__((aligned(8))) attribute_used w04= 0x0004000400040004LL; +static uint64_t __attribute__((aligned(8))) attribute_used w20= 0x0020002000200020LL; +static uint64_t __attribute__((aligned(8))) attribute_used b00= 0x0000000000000000LL; +static uint64_t __attribute__((aligned(8))) attribute_used b01= 0x0101010101010101LL; +static uint64_t __attribute__((aligned(8))) attribute_used b02= 0x0202020202020202LL; +static uint64_t __attribute__((aligned(8))) attribute_used b08= 0x0808080808080808LL; +static uint64_t __attribute__((aligned(8))) attribute_used b80= 0x8080808080808080LL; #endif static uint8_t clip_table[3*256]; @@ -140,65 +140,65 @@ static struct PPFilter filters[]= { - {"hb", "hdeblock", 1, 1, 3, H_DEBLOCK}, - {"vb", "vdeblock", 1, 2, 4, V_DEBLOCK}, -/* {"hr", "rkhdeblock", 1, 1, 3, H_RK1_FILTER}, - {"vr", "rkvdeblock", 1, 2, 4, V_RK1_FILTER},*/ - {"h1", "x1hdeblock", 1, 1, 3, H_X1_FILTER}, - {"v1", "x1vdeblock", 1, 2, 4, V_X1_FILTER}, - {"ha", "ahdeblock", 1, 1, 3, H_A_DEBLOCK}, - {"va", "avdeblock", 1, 2, 4, V_A_DEBLOCK}, - {"dr", "dering", 1, 5, 6, DERING}, - {"al", "autolevels", 0, 1, 2, LEVEL_FIX}, - {"lb", "linblenddeint", 1, 1, 4, LINEAR_BLEND_DEINT_FILTER}, - {"li", "linipoldeint", 1, 1, 4, LINEAR_IPOL_DEINT_FILTER}, - {"ci", "cubicipoldeint", 1, 1, 4, CUBIC_IPOL_DEINT_FILTER}, - {"md", "mediandeint", 1, 1, 4, MEDIAN_DEINT_FILTER}, - {"fd", "ffmpegdeint", 1, 1, 4, FFMPEG_DEINT_FILTER}, - {"l5", "lowpass5", 1, 1, 4, LOWPASS5_DEINT_FILTER}, - {"tn", "tmpnoise", 1, 7, 8, TEMP_NOISE_FILTER}, - {"fq", "forcequant", 1, 0, 0, FORCE_QUANT}, - {NULL, NULL,0,0,0,0} //End Marker + {"hb", "hdeblock", 1, 1, 3, H_DEBLOCK}, + {"vb", "vdeblock", 1, 2, 4, V_DEBLOCK}, +/* {"hr", "rkhdeblock", 1, 1, 3, H_RK1_FILTER}, + {"vr", "rkvdeblock", 1, 2, 4, V_RK1_FILTER},*/ + {"h1", "x1hdeblock", 1, 1, 3, H_X1_FILTER}, + {"v1", "x1vdeblock", 1, 2, 4, V_X1_FILTER}, + {"ha", "ahdeblock", 1, 1, 3, H_A_DEBLOCK}, + {"va", "avdeblock", 1, 2, 4, V_A_DEBLOCK}, + {"dr", "dering", 1, 5, 6, DERING}, + {"al", "autolevels", 0, 1, 2, LEVEL_FIX}, + {"lb", "linblenddeint", 1, 1, 4, LINEAR_BLEND_DEINT_FILTER}, + {"li", "linipoldeint", 1, 1, 4, LINEAR_IPOL_DEINT_FILTER}, + {"ci", "cubicipoldeint", 1, 1, 4, CUBIC_IPOL_DEINT_FILTER}, + {"md", "mediandeint", 1, 1, 4, MEDIAN_DEINT_FILTER}, + {"fd", "ffmpegdeint", 1, 1, 4, FFMPEG_DEINT_FILTER}, + {"l5", "lowpass5", 1, 1, 4, LOWPASS5_DEINT_FILTER}, + {"tn", "tmpnoise", 1, 7, 8, TEMP_NOISE_FILTER}, + {"fq", "forcequant", 1, 0, 0, FORCE_QUANT}, + {NULL, NULL,0,0,0,0} //End Marker }; static char *replaceTable[]= { - "default", "hdeblock:a,vdeblock:a,dering:a", - "de", "hdeblock:a,vdeblock:a,dering:a", - "fast", "x1hdeblock:a,x1vdeblock:a,dering:a", - "fa", "x1hdeblock:a,x1vdeblock:a,dering:a", - "ac", "ha:a:128:7,va:a,dering:a", - NULL //End Marker + "default", "hdeblock:a,vdeblock:a,dering:a", + "de", "hdeblock:a,vdeblock:a,dering:a", + "fast", "x1hdeblock:a,x1vdeblock:a,dering:a", + "fa", "x1hdeblock:a,x1vdeblock:a,dering:a", + "ac", "ha:a:128:7,va:a,dering:a", + NULL //End Marker }; #if defined(ARCH_X86) || defined(ARCH_X86_64) static inline void prefetchnta(void *p) { - asm volatile( "prefetchnta (%0)\n\t" - : : "r" (p) - ); + asm volatile( "prefetchnta (%0)\n\t" + : : "r" (p) + ); } static inline void prefetcht0(void *p) { - asm volatile( "prefetcht0 (%0)\n\t" - : : "r" (p) - ); + asm volatile( "prefetcht0 (%0)\n\t" + : : "r" (p) + ); } static inline void prefetcht1(void *p) { - asm volatile( "prefetcht1 (%0)\n\t" - : : "r" (p) - ); + asm volatile( "prefetcht1 (%0)\n\t" + : : "r" (p) + ); } static inline void prefetcht2(void *p) { - asm volatile( "prefetcht2 (%0)\n\t" - : : "r" (p) - ); + asm volatile( "prefetcht2 (%0)\n\t" + : : "r" (p) + ); } #endif @@ -209,171 +209,171 @@ */ static inline int isHorizDC_C(uint8_t src[], int stride, PPContext *c) { - int numEq= 0; - int y; - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; - const int dcThreshold= dcOffset*2 + 1; + int numEq= 0; + int y; + const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; + const int dcThreshold= dcOffset*2 + 1; - for(y=0; y<BLOCK_SIZE; y++) - { - if(((unsigned)(src[0] - src[1] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[1] - src[2] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[2] - src[3] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[3] - src[4] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[4] - src[5] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[5] - src[6] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[6] - src[7] + dcOffset)) < dcThreshold) numEq++; - src+= stride; - } - return numEq > c->ppMode.flatnessThreshold; + for(y=0; y<BLOCK_SIZE; y++) + { + if(((unsigned)(src[0] - src[1] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[1] - src[2] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[2] - src[3] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[3] - src[4] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[4] - src[5] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[5] - src[6] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[6] - src[7] + dcOffset)) < dcThreshold) numEq++; + src+= stride; + } + return numEq > c->ppMode.flatnessThreshold; } /** * Check if the middle 8x8 Block in the given 8x16 block is flat */ static inline int isVertDC_C(uint8_t src[], int stride, PPContext *c){ - int numEq= 0; - int y; - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; - const int dcThreshold= dcOffset*2 + 1; + int numEq= 0; + int y; + const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; + const int dcThreshold= dcOffset*2 + 1; - src+= stride*4; // src points to begin of the 8x8 Block - for(y=0; y<BLOCK_SIZE-1; y++) - { - if(((unsigned)(src[0] - src[0+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[1] - src[1+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[2] - src[2+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[3] - src[3+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[4] - src[4+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[5] - src[5+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[6] - src[6+stride] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[7] - src[7+stride] + dcOffset)) < dcThreshold) numEq++; - src+= stride; - } - return numEq > c->ppMode.flatnessThreshold; + src+= stride*4; // src points to begin of the 8x8 Block + for(y=0; y<BLOCK_SIZE-1; y++) + { + if(((unsigned)(src[0] - src[0+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[1] - src[1+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[2] - src[2+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[3] - src[3+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[4] - src[4+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[5] - src[5+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[6] - src[6+stride] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[7] - src[7+stride] + dcOffset)) < dcThreshold) numEq++; + src+= stride; + } + return numEq > c->ppMode.flatnessThreshold; } static inline int isHorizMinMaxOk_C(uint8_t src[], int stride, int QP) { - int i; + int i; #if 1 - for(i=0; i<2; i++){ - if((unsigned)(src[0] - src[5] + 2*QP) > 4*QP) return 0; - src += stride; - if((unsigned)(src[2] - src[7] + 2*QP) > 4*QP) return 0; - src += stride; - if((unsigned)(src[4] - src[1] + 2*QP) > 4*QP) return 0; - src += stride; - if((unsigned)(src[6] - src[3] + 2*QP) > 4*QP) return 0; - src += stride; - } + for(i=0; i<2; i++){ + if((unsigned)(src[0] - src[5] + 2*QP) > 4*QP) return 0; + src += stride; + if((unsigned)(src[2] - src[7] + 2*QP) > 4*QP) return 0; + src += stride; + if((unsigned)(src[4] - src[1] + 2*QP) > 4*QP) return 0; + src += stride; + if((unsigned)(src[6] - src[3] + 2*QP) > 4*QP) return 0; + src += stride; + } #else - for(i=0; i<8; i++){ - if((unsigned)(src[0] - src[7] + 2*QP) > 4*QP) return 0; - src += stride; - } + for(i=0; i<8; i++){ + if((unsigned)(src[0] - src[7] + 2*QP) > 4*QP) return 0; + src += stride; + } #endif - return 1; + return 1; } static inline int isVertMinMaxOk_C(uint8_t src[], int stride, int QP) { #if 1 #if 1 - int x; - src+= stride*4; - for(x=0; x<BLOCK_SIZE; x+=4) - { - if((unsigned)(src[ x + 0*stride] - src[ x + 5*stride] + 2*QP) > 4*QP) return 0; - if((unsigned)(src[1+x + 2*stride] - src[1+x + 7*stride] + 2*QP) > 4*QP) return 0; - if((unsigned)(src[2+x + 4*stride] - src[2+x + 1*stride] + 2*QP) > 4*QP) return 0; - if((unsigned)(src[3+x + 6*stride] - src[3+x + 3*stride] + 2*QP) > 4*QP) return 0; - } + int x; + src+= stride*4; + for(x=0; x<BLOCK_SIZE; x+=4) + { + if((unsigned)(src[ x + 0*stride] - src[ x + 5*stride] + 2*QP) > 4*QP) return 0; + if((unsigned)(src[1+x + 2*stride] - src[1+x + 7*stride] + 2*QP) > 4*QP) return 0; + if((unsigned)(src[2+x + 4*stride] - src[2+x + 1*stride] + 2*QP) > 4*QP) return 0; + if((unsigned)(src[3+x + 6*stride] - src[3+x + 3*stride] + 2*QP) > 4*QP) return 0; + } #else - int x; - src+= stride*3; - for(x=0; x<BLOCK_SIZE; x++) - { - if((unsigned)(src[x + stride] - src[x + (stride<<3)] + 2*QP) > 4*QP) return 0; - } + int x; + src+= stride*3; + for(x=0; x<BLOCK_SIZE; x++) + { + if((unsigned)(src[x + stride] - src[x + (stride<<3)] + 2*QP) > 4*QP) return 0; + } #endif - return 1; + return 1; #else - int x; - src+= stride*4; - for(x=0; x<BLOCK_SIZE; x++) - { - int min=255; - int max=0; - int y; - for(y=0; y<8; y++){ - int v= src[x + y*stride]; - if(v>max) max=v; - if(v<min) min=v; - } - if(max-min > 2*QP) return 0; - } - return 1; + int x; + src+= stride*4; + for(x=0; x<BLOCK_SIZE; x++) + { + int min=255; + int max=0; + int y; + for(y=0; y<8; y++){ + int v= src[x + y*stride]; + if(v>max) max=v; + if(v<min) min=v; + } + if(max-min > 2*QP) return 0; + } + return 1; #endif } static inline int horizClassify_C(uint8_t src[], int stride, PPContext *c){ - if( isHorizDC_C(src, stride, c) ){ - if( isHorizMinMaxOk_C(src, stride, c->QP) ) - return 1; - else - return 0; - }else{ - return 2; - } + if( isHorizDC_C(src, stride, c) ){ + if( isHorizMinMaxOk_C(src, stride, c->QP) ) + return 1; + else + return 0; + }else{ + return 2; + } } static inline int vertClassify_C(uint8_t src[], int stride, PPContext *c){ - if( isVertDC_C(src, stride, c) ){ - if( isVertMinMaxOk_C(src, stride, c->QP) ) - return 1; - else - return 0; - }else{ - return 2; - } + if( isVertDC_C(src, stride, c) ){ + if( isVertMinMaxOk_C(src, stride, c->QP) ) + return 1; + else + return 0; + }else{ + return 2; + } } static inline void doHorizDefFilter_C(uint8_t dst[], int stride, PPContext *c) { - int y; - for(y=0; y<BLOCK_SIZE; y++) - { - const int middleEnergy= 5*(dst[4] - dst[3]) + 2*(dst[2] - dst[5]); + int y; + for(y=0; y<BLOCK_SIZE; y++) + { + const int middleEnergy= 5*(dst[4] - dst[3]) + 2*(dst[2] - dst[5]); - if(ABS(middleEnergy) < 8*c->QP) - { - const int q=(dst[3] - dst[4])/2; - const int leftEnergy= 5*(dst[2] - dst[1]) + 2*(dst[0] - dst[3]); - const int rightEnergy= 5*(dst[6] - dst[5]) + 2*(dst[4] - dst[7]); + if(ABS(middleEnergy) < 8*c->QP) + { + const int q=(dst[3] - dst[4])/2; + const int leftEnergy= 5*(dst[2] - dst[1]) + 2*(dst[0] - dst[3]); + const int rightEnergy= 5*(dst[6] - dst[5]) + 2*(dst[4] - dst[7]); - int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); - d= MAX(d, 0); + int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); + d= MAX(d, 0); - d= (5*d + 32) >> 6; - d*= SIGN(-middleEnergy); + d= (5*d + 32) >> 6; + d*= SIGN(-middleEnergy); - if(q>0) - { - d= d<0 ? 0 : d; - d= d>q ? q : d; - } - else - { - d= d>0 ? 0 : d; - d= d<q ? q : d; - } + if(q>0) + { + d= d<0 ? 0 : d; + d= d>q ? q : d; + } + else + { + d= d>0 ? 0 : d; + d= d<q ? q : d; + } - dst[3]-= d; - dst[4]+= d; - } - dst+= stride; - } + dst[3]-= d; + dst[4]+= d; + } + dst+= stride; + } } /** @@ -382,35 +382,35 @@ */ static inline void doHorizLowPass_C(uint8_t dst[], int stride, PPContext *c) { - int y; - for(y=0; y<BLOCK_SIZE; y++) - { - const int first= ABS(dst[-1] - dst[0]) < c->QP ? dst[-1] : dst[0]; - const int last= ABS(dst[8] - dst[7]) < c->QP ? dst[8] : dst[7]; + int y; + for(y=0; y<BLOCK_SIZE; y++) + { + const int first= ABS(dst[-1] - dst[0]) < c->QP ? dst[-1] : dst[0]; + const int last= ABS(dst[8] - dst[7]) < c->QP ? dst[8] : dst[7]; - int sums[10]; - sums[0] = 4*first + dst[0] + dst[1] + dst[2] + 4; - sums[1] = sums[0] - first + dst[3]; - sums[2] = sums[1] - first + dst[4]; - sums[3] = sums[2] - first + dst[5]; - sums[4] = sums[3] - first + dst[6]; - sums[5] = sums[4] - dst[0] + dst[7]; - sums[6] = sums[5] - dst[1] + last; - sums[7] = sums[6] - dst[2] + last; - sums[8] = sums[7] - dst[3] + last; - sums[9] = sums[8] - dst[4] + last; + int sums[10]; + sums[0] = 4*first + dst[0] + dst[1] + dst[2] + 4; + sums[1] = sums[0] - first + dst[3]; + sums[2] = sums[1] - first + dst[4]; + sums[3] = sums[2] - first + dst[5]; + sums[4] = sums[3] - first + dst[6]; + sums[5] = sums[4] - dst[0] + dst[7]; + sums[6] = sums[5] - dst[1] + last; + sums[7] = sums[6] - dst[2] + last; + sums[8] = sums[7] - dst[3] + last; + sums[9] = sums[8] - dst[4] + last; - dst[0]= (sums[0] + sums[2] + 2*dst[0])>>4; - dst[1]= (sums[1] + sums[3] + 2*dst[1])>>4; - dst[2]= (sums[2] + sums[4] + 2*dst[2])>>4; - dst[3]= (sums[3] + sums[5] + 2*dst[3])>>4; - dst[4]= (sums[4] + sums[6] + 2*dst[4])>>4; - dst[5]= (sums[5] + sums[7] + 2*dst[5])>>4; - dst[6]= (sums[6] + sums[8] + 2*dst[6])>>4; - dst[7]= (sums[7] + sums[9] + 2*dst[7])>>4; + dst[0]= (sums[0] + sums[2] + 2*dst[0])>>4; + dst[1]= (sums[1] + sums[3] + 2*dst[1])>>4; + dst[2]= (sums[2] + sums[4] + 2*dst[2])>>4; + dst[3]= (sums[3] + sums[5] + 2*dst[3])>>4; + dst[4]= (sums[4] + sums[6] + 2*dst[4])>>4; + dst[5]= (sums[5] + sums[7] + 2*dst[5])>>4; + dst[6]= (sums[6] + sums[8] + 2*dst[6])>>4; + dst[7]= (sums[7] + sums[9] + 2*dst[7])>>4; - dst+= stride; - } + dst+= stride; + } } /** @@ -423,161 +423,161 @@ */ static inline void horizX1Filter(uint8_t *src, int stride, int QP) { - int y; - static uint64_t *lut= NULL; - if(lut==NULL) - { - int i; - lut= (uint64_t*)memalign(8, 256*8); - for(i=0; i<256; i++) - { - int v= i < 128 ? 2*i : 2*(i-256); + int y; + static uint64_t *lut= NULL; + if(lut==NULL) + { + int i; + lut= (uint64_t*)memalign(8, 256*8); + for(i=0; i<256; i++) + { + int v= i < 128 ? 2*i : 2*(i-256); /* //Simulate 112242211 9-Tap filter - uint64_t a= (v/16) & 0xFF; - uint64_t b= (v/8) & 0xFF; - uint64_t c= (v/4) & 0xFF; - uint64_t d= (3*v/8) & 0xFF; + uint64_t a= (v/16) & 0xFF; + uint64_t b= (v/8) & 0xFF; + uint64_t c= (v/4) & 0xFF; + uint64_t d= (3*v/8) & 0xFF; */ //Simulate piecewise linear interpolation - uint64_t a= (v/16) & 0xFF; - uint64_t b= (v*3/16) & 0xFF; - uint64_t c= (v*5/16) & 0xFF; - uint64_t d= (7*v/16) & 0xFF; - uint64_t A= (0x100 - a)&0xFF; - uint64_t B= (0x100 - b)&0xFF; - uint64_t C= (0x100 - c)&0xFF; - uint64_t D= (0x100 - c)&0xFF; + uint64_t a= (v/16) & 0xFF; + uint64_t b= (v*3/16) & 0xFF; + uint64_t c= (v*5/16) & 0xFF; + uint64_t d= (7*v/16) & 0xFF; + uint64_t A= (0x100 - a)&0xFF; + uint64_t B= (0x100 - b)&0xFF; + uint64_t C= (0x100 - c)&0xFF; + uint64_t D= (0x100 - c)&0xFF; - lut[i] = (a<<56) | (b<<48) | (c<<40) | (d<<32) | - (D<<24) | (C<<16) | (B<<8) | (A); - //lut[i] = (v<<32) | (v<<24); - } - } + lut[i] = (a<<56) | (b<<48) | (c<<40) | (d<<32) | + (D<<24) | (C<<16) | (B<<8) | (A); + //lut[i] = (v<<32) | (v<<24); + } + } - for(y=0; y<BLOCK_SIZE; y++) - { - int a= src[1] - src[2]; - int b= src[3] - src[4]; - int c= src[5] - src[6]; + for(y=0; y<BLOCK_SIZE; y++) + { + int a= src[1] - src[2]; + int b= src[3] - src[4]; + int c= src[5] - src[6]; - int d= MAX(ABS(b) - (ABS(a) + ABS(c))/2, 0); + int d= MAX(ABS(b) - (ABS(a) + ABS(c))/2, 0); - if(d < QP) - { - int v = d * SIGN(-b); + if(d < QP) + { + int v = d * SIGN(-b); - src[1] +=v/8; - src[2] +=v/4; - src[3] +=3*v/8; - src[4] -=3*v/8; - src[5] -=v/4; - src[6] -=v/8; + src[1] +=v/8; + src[2] +=v/4; + src[3] +=3*v/8; + src[4] -=3*v/8; + src[5] -=v/4; + src[6] -=v/8; - } - src+=stride; - } + } + src+=stride; + } } /** * accurate deblock filter */ static always_inline void do_a_deblock_C(uint8_t *src, int step, int stride, PPContext *c){ - int y; - const int QP= c->QP; - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; - const int dcThreshold= dcOffset*2 + 1; + int y; + const int QP= c->QP; + const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1; + const int dcThreshold= dcOffset*2 + 1; //START_TIMER - src+= step*4; // src points to begin of the 8x8 Block - for(y=0; y<8; y++){ - int numEq= 0; + src+= step*4; // src points to begin of the 8x8 Block + for(y=0; y<8; y++){ + int numEq= 0; - if(((unsigned)(src[-1*step] - src[0*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 0*step] - src[1*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 1*step] - src[2*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 2*step] - src[3*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 3*step] - src[4*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 4*step] - src[5*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 5*step] - src[6*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 6*step] - src[7*step] + dcOffset)) < dcThreshold) numEq++; - if(((unsigned)(src[ 7*step] - src[8*step] + dcOffset)) < dcThreshold) numEq++; - if(numEq > c->ppMode.flatnessThreshold){ - int min, max, x; + if(((unsigned)(src[-1*step] - src[0*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 0*step] - src[1*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 1*step] - src[2*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 2*step] - src[3*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 3*step] - src[4*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 4*step] - src[5*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 5*step] - src[6*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 6*step] - src[7*step] + dcOffset)) < dcThreshold) numEq++; + if(((unsigned)(src[ 7*step] - src[8*step] + dcOffset)) < dcThreshold) numEq++; + if(numEq > c->ppMode.flatnessThreshold){ + int min, max, x; - if(src[0] > src[step]){ - max= src[0]; - min= src[step]; - }else{ - max= src[step]; - min= src[0]; - } - for(x=2; x<8; x+=2){ - if(src[x*step] > src[(x+1)*step]){ - if(src[x *step] > max) max= src[ x *step]; - if(src[(x+1)*step] < min) min= src[(x+1)*step]; - }else{ - if(src[(x+1)*step] > max) max= src[(x+1)*step]; - if(src[ x *step] < min) min= src[ x *step]; - } - } - if(max-min < 2*QP){ - const int first= ABS(src[-1*step] - src[0]) < QP ? src[-1*step] : src[0]; - const int last= ABS(src[8*step] - src[7*step]) < QP ? src[8*step] : src[7*step]; + if(src[0] > src[step]){ + max= src[0]; + min= src[step]; + }else{ + max= src[step]; + min= src[0]; + } + for(x=2; x<8; x+=2){ + if(src[x*step] > src[(x+1)*step]){ + if(src[x *step] > max) max= src[ x *step]; + if(src[(x+1)*step] < min) min= src[(x+1)*step]; + }else{ + if(src[(x+1)*step] > max) max= src[(x+1)*step]; + if(src[ x *step] < min) min= src[ x *step]; + } + } + if(max-min < 2*QP){ + const int first= ABS(src[-1*step] - src[0]) < QP ? src[-1*step] : src[0]; + const int last= ABS(src[8*step] - src[7*step]) < QP ? src[8*step] : src[7*step]; - int sums[10]; - sums[0] = 4*first + src[0*step] + src[1*step] + src[2*step] + 4; - sums[1] = sums[0] - first + src[3*step]; - sums[2] = sums[1] - first + src[4*step]; - sums[3] = sums[2] - first + src[5*step]; - sums[4] = sums[3] - first + src[6*step]; - sums[5] = sums[4] - src[0*step] + src[7*step]; - sums[6] = sums[5] - src[1*step] + last; - sums[7] = sums[6] - src[2*step] + last; - sums[8] = sums[7] - src[3*step] + last; - sums[9] = sums[8] - src[4*step] + last; + int sums[10]; + sums[0] = 4*first + src[0*step] + src[1*step] + src[2*step] + 4; + sums[1] = sums[0] - first + src[3*step]; + sums[2] = sums[1] - first + src[4*step]; + sums[3] = sums[2] - first + src[5*step]; + sums[4] = sums[3] - first + src[6*step]; + sums[5] = sums[4] - src[0*step] + src[7*step]; + sums[6] = sums[5] - src[1*step] + last; + sums[7] = sums[6] - src[2*step] + last; + sums[8] = sums[7] - src[3*step] + last; + sums[9] = sums[8] - src[4*step] + last; - src[0*step]= (sums[0] + sums[2] + 2*src[0*step])>>4; - src[1*step]= (sums[1] + sums[3] + 2*src[1*step])>>4; - src[2*step]= (sums[2] + sums[4] + 2*src[2*step])>>4; - src[3*step]= (sums[3] + sums[5] + 2*src[3*step])>>4; - src[4*step]= (sums[4] + sums[6] + 2*src[4*step])>>4; - src[5*step]= (sums[5] + sums[7] + 2*src[5*step])>>4; - src[6*step]= (sums[6] + sums[8] + 2*src[6*step])>>4; - src[7*step]= (sums[7] + sums[9] + 2*src[7*step])>>4; - } - }else{ - const int middleEnergy= 5*(src[4*step] - src[3*step]) + 2*(src[2*step] - src[5*step]); + src[0*step]= (sums[0] + sums[2] + 2*src[0*step])>>4; + src[1*step]= (sums[1] + sums[3] + 2*src[1*step])>>4; + src[2*step]= (sums[2] + sums[4] + 2*src[2*step])>>4; + src[3*step]= (sums[3] + sums[5] + 2*src[3*step])>>4; + src[4*step]= (sums[4] + sums[6] + 2*src[4*step])>>4; + src[5*step]= (sums[5] + sums[7] + 2*src[5*step])>>4; + src[6*step]= (sums[6] + sums[8] + 2*src[6*step])>>4; + src[7*step]= (sums[7] + sums[9] + 2*src[7*step])>>4; + } + }else{ + const int middleEnergy= 5*(src[4*step] - src[3*step]) + 2*(src[2*step] - src[5*step]); - if(ABS(middleEnergy) < 8*QP) - { - const int q=(src[3*step] - src[4*step])/2; - const int leftEnergy= 5*(src[2*step] - src[1*step]) + 2*(src[0*step] - src[3*step]); - const int rightEnergy= 5*(src[6*step] - src[5*step]) + 2*(src[4*step] - src[7*step]); + if(ABS(middleEnergy) < 8*QP) + { + const int q=(src[3*step] - src[4*step])/2; + const int leftEnergy= 5*(src[2*step] - src[1*step]) + 2*(src[0*step] - src[3*step]); + const int rightEnergy= 5*(src[6*step] - src[5*step]) + 2*(src[4*step] - src[7*step]); - int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); - d= MAX(d, 0); + int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); + d= MAX(d, 0); - d= (5*d + 32) >> 6; - d*= SIGN(-middleEnergy); + d= (5*d + 32) >> 6; + d*= SIGN(-middleEnergy); - if(q>0) - { - d= d<0 ? 0 : d; - d= d>q ? q : d; - } - else - { - d= d>0 ? 0 : d; - d= d<q ? q : d; - } + if(q>0) + { + d= d<0 ? 0 : d; + d= d>q ? q : d; + } + else + { + d= d>0 ? 0 : d; + d= d<q ? q : d; + } - src[3*step]-= d; - src[4*step]+= d; - } - } + src[3*step]-= d; + src[4*step]+= d; + } + } - src += stride; - } + src += stride; + } /*if(step==16){ STOP_TIMER("step16") }else{ @@ -668,290 +668,290 @@ // minor note: the HAVE_xyz is messed up after that line so dont use it static inline void postProcess(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height, - QP_STORE_T QPs[], int QPStride, int isColor, pp_mode_t *vm, pp_context_t *vc) + QP_STORE_T QPs[], int QPStride, int isColor, pp_mode_t *vm, pp_context_t *vc) { - PPContext *c= (PPContext *)vc; - PPMode *ppMode= (PPMode *)vm; - c->ppMode= *ppMode; //FIXME + PPContext *c= (PPContext *)vc; + PPMode *ppMode= (PPMode *)vm; + c->ppMode= *ppMode; //FIXME - // useing ifs here as they are faster than function pointers allthough the - // difference wouldnt be messureable here but its much better because - // someone might exchange the cpu whithout restarting mplayer ;) + // useing ifs here as they are faster than function pointers allthough the + // difference wouldnt be messureable here but its much better because + // someone might exchange the cpu whithout restarting mplayer ;) #ifdef RUNTIME_CPUDETECT #if defined(ARCH_X86) || defined(ARCH_X86_64) - // ordered per speed fasterst first - if(c->cpuCaps & PP_CPU_CAPS_MMX2) - postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); - else if(c->cpuCaps & PP_CPU_CAPS_3DNOW) - postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); - else if(c->cpuCaps & PP_CPU_CAPS_MMX) - postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); - else - postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + // ordered per speed fasterst first + if(c->cpuCaps & PP_CPU_CAPS_MMX2) + postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + else if(c->cpuCaps & PP_CPU_CAPS_3DNOW) + postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + else if(c->cpuCaps & PP_CPU_CAPS_MMX) + postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + else + postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #else #ifdef ARCH_POWERPC #ifdef HAVE_ALTIVEC if(c->cpuCaps & PP_CPU_CAPS_ALTIVEC) - postProcess_altivec(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_altivec(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); else #endif #endif - postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #endif #else //RUNTIME_CPUDETECT #ifdef HAVE_MMX2 - postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #elif defined (HAVE_3DNOW) - postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #elif defined (HAVE_MMX) - postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #elif defined (HAVE_ALTIVEC) - postProcess_altivec(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_altivec(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #else - postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); + postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c); #endif #endif //!RUNTIME_CPUDETECT } //static void postProcess(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height, -// QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode); +// QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode); /* -pp Command line Help */ char *pp_help= "Available postprocessing filters:\n" -"Filters Options\n" -"short long name short long option Description\n" -"* * a autoq CPU power dependent enabler\n" -" c chrom chrominance filtering enabled\n" -" y nochrom chrominance filtering disabled\n" -" n noluma luma filtering disabled\n" -"hb hdeblock (2 threshold) horizontal deblocking filter\n" -" 1. difference factor: default=32, higher -> more deblocking\n" -" 2. flatness threshold: default=39, lower -> more deblocking\n" -" the h & v deblocking filters share these\n" -" so you can't set different thresholds for h / v\n" -"vb vdeblock (2 threshold) vertical deblocking filter\n" -"ha hadeblock (2 threshold) horizontal deblocking filter\n" -"va vadeblock (2 threshold) vertical deblocking filter\n" -"h1 x1hdeblock experimental h deblock filter 1\n" -"v1 x1vdeblock experimental v deblock filter 1\n" -"dr dering deringing filter\n" -"al autolevels automatic brightness / contrast\n" -" f fullyrange stretch luminance to (0..255)\n" -"lb linblenddeint linear blend deinterlacer\n" -"li linipoldeint linear interpolating deinterlace\n" -"ci cubicipoldeint cubic interpolating deinterlacer\n" -"md mediandeint median deinterlacer\n" -"fd ffmpegdeint ffmpeg deinterlacer\n" -"l5 lowpass5 FIR lowpass deinterlacer\n" -"de default hb:a,vb:a,dr:a\n" -"fa fast h1:a,v1:a,dr:a\n" -"ac ha:a:128:7,va:a,dr:a\n" -"tn tmpnoise (3 threshold) temporal noise reducer\n" -" 1. <= 2. <= 3. larger -> stronger filtering\n" -"fq forceQuant <quantizer> force quantizer\n" +"Filters Options\n" +"short long name short long option Description\n" +"* * a autoq CPU power dependent enabler\n" +" c chrom chrominance filtering enabled\n" +" y nochrom chrominance filtering disabled\n" +" n noluma luma filtering disabled\n" +"hb hdeblock (2 threshold) horizontal deblocking filter\n" +" 1. difference factor: default=32, higher -> more deblocking\n" +" 2. flatness threshold: default=39, lower -> more deblocking\n" +" the h & v deblocking filters share these\n" +" so you can't set different thresholds for h / v\n" +"vb vdeblock (2 threshold) vertical deblocking filter\n" +"ha hadeblock (2 threshold) horizontal deblocking filter\n" +"va vadeblock (2 threshold) vertical deblocking filter\n" +"h1 x1hdeblock experimental h deblock filter 1\n" +"v1 x1vdeblock experimental v deblock filter 1\n" +"dr dering deringing filter\n" +"al autolevels automatic brightness / contrast\n" +" f fullyrange stretch luminance to (0..255)\n" +"lb linblenddeint linear blend deinterlacer\n" +"li linipoldeint linear interpolating deinterlace\n" +"ci cubicipoldeint cubic interpolating deinterlacer\n" +"md mediandeint median deinterlacer\n" +"fd ffmpegdeint ffmpeg deinterlacer\n" +"l5 lowpass5 FIR lowpass deinterlacer\n" +"de default hb:a,vb:a,dr:a\n" +"fa fast h1:a,v1:a,dr:a\n" +"ac ha:a:128:7,va:a,dr:a\n" +"tn tmpnoise (3 threshold) temporal noise reducer\n" +" 1. <= 2. <= 3. larger -> stronger filtering\n" +"fq forceQuant <quantizer> force quantizer\n" "Usage:\n" "<filterName>[:<option>[:<option>...]][[,|/][-]<filterName>[:<option>...]]...\n" "long form example:\n" -"vdeblock:autoq/hdeblock:autoq/linblenddeint default,-vdeblock\n" +"vdeblock:autoq/hdeblock:autoq/linblenddeint default,-vdeblock\n" "short form example:\n" -"vb:a/hb:a/lb de,-vb\n" +"vb:a/hb:a/lb de,-vb\n" "more examples:\n" "tn:64:128:256\n" ; pp_mode_t *pp_get_mode_by_name_and_quality(char *name, int quality) { - char temp[GET_MODE_BUFFER_SIZE]; - char *p= temp; - char *filterDelimiters= ",/"; - char *optionDelimiters= ":"; - struct PPMode *ppMode; - char *filterToken; + char temp[GET_MODE_BUFFER_SIZE]; + char *p= temp; + char *filterDelimiters= ",/"; + char *optionDelimiters= ":"; + struct PPMode *ppMode; + char *filterToken; - ppMode= memalign(8, sizeof(PPMode)); + ppMode= memalign(8, sizeof(PPMode)); - ppMode->lumMode= 0; - ppMode->chromMode= 0; - ppMode->maxTmpNoise[0]= 700; - ppMode->maxTmpNoise[1]= 1500; - ppMode->maxTmpNoise[2]= 3000; - ppMode->maxAllowedY= 234; - ppMode->minAllowedY= 16; - ppMode->baseDcDiff= 256/8; - ppMode->flatnessThreshold= 56-16-1; - ppMode->maxClippedThreshold= 0.01; - ppMode->error=0; + ppMode->lumMode= 0; + ppMode->chromMode= 0; + ppMode->maxTmpNoise[0]= 700; + ppMode->maxTmpNoise[1]= 1500; + ppMode->maxTmpNoise[2]= 3000; + ppMode->maxAllowedY= 234; + ppMode->minAllowedY= 16; + ppMode->baseDcDiff= 256/8; + ppMode->flatnessThreshold= 56-16-1; + ppMode->maxClippedThreshold= 0.01; + ppMode->error=0; - strncpy(temp, name, GET_MODE_BUFFER_SIZE); + strncpy(temp, name, GET_MODE_BUFFER_SIZE); - if(verbose>1) printf("pp: %s\n", name); + if(verbose>1) printf("pp: %s\n", name); - for(;;){ - char *filterName; - int q= 1000000; //PP_QUALITY_MAX; - int chrom=-1; - int luma=-1; - char *option; - char *options[OPTIONS_ARRAY_SIZE]; - int i; - int filterNameOk=0; - int numOfUnknownOptions=0; - int enable=1; //does the user want us to enabled or disabled the filter + for(;;){ + char *filterName; + int q= 1000000; //PP_QUALITY_MAX; + int chrom=-1; + int luma=-1; + char *option; + char *options[OPTIONS_ARRAY_SIZE]; + int i; + int filterNameOk=0; + int numOfUnknownOptions=0; + int enable=1; //does the user want us to enabled or disabled the filter - filterToken= strtok(p, filterDelimiters); - if(filterToken == NULL) break; - p+= strlen(filterToken) + 1; // p points to next filterToken - filterName= strtok(filterToken, optionDelimiters); - if(verbose>1) printf("pp: %s::%s\n", filterToken, filterName); + filterToken= strtok(p, filterDelimiters); + if(filterToken == NULL) break; + p+= strlen(filterToken) + 1; // p points to next filterToken + filterName= strtok(filterToken, optionDelimiters); + if(verbose>1) printf("pp: %s::%s\n", filterToken, filterName); - if(*filterName == '-') - { - enable=0; - filterName++; - } + if(*filterName == '-') + { + enable=0; + filterName++; + } - for(;;){ //for all options - option= strtok(NULL, optionDelimiters); - if(option == NULL) break; + for(;;){ //for all options + option= strtok(NULL, optionDelimiters); + if(option == NULL) break; - if(verbose>1) printf("pp: option: %s\n", option); - if(!strcmp("autoq", option) || !strcmp("a", option)) q= quality; - else if(!strcmp("nochrom", option) || !strcmp("y", option)) chrom=0; - else if(!strcmp("chrom", option) || !strcmp("c", option)) chrom=1; - else if(!strcmp("noluma", option) || !strcmp("n", option)) luma=0; - else - { - options[numOfUnknownOptions] = option; - numOfUnknownOptions++; - } - if(numOfUnknownOptions >= OPTIONS_ARRAY_SIZE-1) break; - } - options[numOfUnknownOptions] = NULL; + if(verbose>1) printf("pp: option: %s\n", option); + if(!strcmp("autoq", option) || !strcmp("a", option)) q= quality; + else if(!strcmp("nochrom", option) || !strcmp("y", option)) chrom=0; + else if(!strcmp("chrom", option) || !strcmp("c", option)) chrom=1; + else if(!strcmp("noluma", option) || !strcmp("n", option)) luma=0; + else + { + options[numOfUnknownOptions] = option; + numOfUnknownOptions++; + } + if(numOfUnknownOptions >= OPTIONS_ARRAY_SIZE-1) break; + } + options[numOfUnknownOptions] = NULL; - /* replace stuff from the replace Table */ - for(i=0; replaceTable[2*i]!=NULL; i++) - { - if(!strcmp(replaceTable[2*i], filterName)) - { - int newlen= strlen(replaceTable[2*i + 1]); - int plen; - int spaceLeft; + /* replace stuff from the replace Table */ + for(i=0; replaceTable[2*i]!=NULL; i++) + { + if(!strcmp(replaceTable[2*i], filterName)) + { + int newlen= strlen(replaceTable[2*i + 1]); + int plen; + int spaceLeft; - if(p==NULL) p= temp, *p=0; //last filter - else p--, *p=','; //not last filter + if(p==NULL) p= temp, *p=0; //last filter + else p--, *p=','; //not last filter - plen= strlen(p); - spaceLeft= p - temp + plen; - if(spaceLeft + newlen >= GET_MODE_BUFFER_SIZE) - { - ppMode->error++; - break; - } - memmove(p + newlen, p, plen+1); - memcpy(p, replaceTable[2*i + 1], newlen); - filterNameOk=1; - } - } + plen= strlen(p); + spaceLeft= p - temp + plen; + if(spaceLeft + newlen >= GET_MODE_BUFFER_SIZE) + { + ppMode->error++; + break; + } + memmove(p + newlen, p, plen+1); + memcpy(p, replaceTable[2*i + 1], newlen); + filterNameOk=1; + } + } - for(i=0; filters[i].shortName!=NULL; i++) - { -// printf("Compareing %s, %s, %s\n", filters[i].shortName,filters[i].longName, filterName); - if( !strcmp(filters[i].longName, filterName) - || !strcmp(filters[i].shortName, filterName)) - { - ppMode->lumMode &= ~filters[i].mask; - ppMode->chromMode &= ~filters[i].mask; + for(i=0; filters[i].shortName!=NULL; i++) + { +// printf("Compareing %s, %s, %s\n", filters[i].shortName,filters[i].longName, filterName); + if( !strcmp(filters[i].longName, filterName) + || !strcmp(filters[i].shortName, filterName)) + { + ppMode->lumMode &= ~filters[i].mask; + ppMode->chromMode &= ~filters[i].mask; - filterNameOk=1; - if(!enable) break; // user wants to disable it + filterNameOk=1; + if(!enable) break; // user wants to disable it - if(q >= filters[i].minLumQuality && luma) - ppMode->lumMode|= filters[i].mask; - if(chrom==1 || (chrom==-1 && filters[i].chromDefault)) - if(q >= filters[i].minChromQuality) - ppMode->chromMode|= filters[i].mask; + if(q >= filters[i].minLumQuality && luma) + ppMode->lumMode|= filters[i].mask; + if(chrom==1 || (chrom==-1 && filters[i].chromDefault)) + if(q >= filters[i].minChromQuality) + ppMode->chromMode|= filters[i].mask; - if(filters[i].mask == LEVEL_FIX) - { - int o; - ppMode->minAllowedY= 16; - ppMode->maxAllowedY= 234; - for(o=0; options[o]!=NULL; o++) - { - if( !strcmp(options[o],"fullyrange") - ||!strcmp(options[o],"f")) - { - ppMode->minAllowedY= 0; - ppMode->maxAllowedY= 255; - numOfUnknownOptions--; - } - } - } - else if(filters[i].mask == TEMP_NOISE_FILTER) - { - int o; - int numOfNoises=0; + if(filters[i].mask == LEVEL_FIX) + { + int o; + ppMode->minAllowedY= 16; + ppMode->maxAllowedY= 234; + for(o=0; options[o]!=NULL; o++) + { + if( !strcmp(options[o],"fullyrange") + ||!strcmp(options[o],"f")) + { + ppMode->minAllowedY= 0; + ppMode->maxAllowedY= 255; + numOfUnknownOptions--; + } + } + } + else if(filters[i].mask == TEMP_NOISE_FILTER) + { + int o; + int numOfNoises=0; - for(o=0; options[o]!=NULL; o++) - { - char *tail; - ppMode->maxTmpNoise[numOfNoises]= - strtol(options[o], &tail, 0); - if(tail!=options[o]) - { - numOfNoises++; - numOfUnknownOptions--; - if(numOfNoises >= 3) break; - } - } - } - else if(filters[i].mask == V_DEBLOCK || filters[i].mask == H_DEBLOCK - || filters[i].mask == V_A_DEBLOCK || filters[i].mask == H_A_DEBLOCK) - { - int o; + for(o=0; options[o]!=NULL; o++) + { + char *tail; + ppMode->maxTmpNoise[numOfNoises]= + strtol(options[o], &tail, 0); + if(tail!=options[o]) + { + numOfNoises++; + numOfUnknownOptions--; + if(numOfNoises >= 3) break; + } + } + } + else if(filters[i].mask == V_DEBLOCK || filters[i].mask == H_DEBLOCK + || filters[i].mask == V_A_DEBLOCK || filters[i].mask == H_A_DEBLOCK) + { + int o; - for(o=0; options[o]!=NULL && o<2; o++) - { - char *tail; - int val= strtol(options[o], &tail, 0); - if(tail==options[o]) break; + for(o=0; options[o]!=NULL && o<2; o++) + { + char *tail; + int val= strtol(options[o], &tail, 0); + if(tail==options[o]) break; - numOfUnknownOptions--; - if(o==0) ppMode->baseDcDiff= val; - else ppMode->flatnessThreshold= val; - } - } - else if(filters[i].mask == FORCE_QUANT) - { - int o; - ppMode->forcedQuant= 15; + numOfUnknownOptions--; + if(o==0) ppMode->baseDcDiff= val; + else ppMode->flatnessThreshold= val; + } + } + else if(filters[i].mask == FORCE_QUANT) + { + int o; + ppMode->forcedQuant= 15; - for(o=0; options[o]!=NULL && o<1; o++) - { - char *tail; - int val= strtol(options[o], &tail, 0); - if(tail==options[o]) break; + for(o=0; options[o]!=NULL && o<1; o++) + { + char *tail; + int val= strtol(options[o], &tail, 0); + if(tail==options[o]) break; - numOfUnknownOptions--; - ppMode->forcedQuant= val; - } - } - } - } - if(!filterNameOk) ppMode->error++; - ppMode->error += numOfUnknownOptions; - } + numOfUnknownOptions--; + ppMode->forcedQuant= val; + } + } + } + } + if(!filterNameOk) ppMode->error++; + ppMode->error += numOfUnknownOptions; + } - if(verbose>1) printf("pp: lumMode=%X, chromMode=%X\n", ppMode->lumMode, ppMode->chromMode); - if(ppMode->error) - { - fprintf(stderr, "%d errors in postprocess string \"%s\"\n", ppMode->error, name); - free(ppMode); - return NULL; - } - return ppMode; + if(verbose>1) printf("pp: lumMode=%X, chromMode=%X\n", ppMode->lumMode, ppMode->chromMode); + if(ppMode->error) + { + fprintf(stderr, "%d errors in postprocess string \"%s\"\n", ppMode->error, name); + free(ppMode); + return NULL; + } + return ppMode; } void pp_free_mode(pp_mode_t *mode){ @@ -959,199 +959,199 @@ } static void reallocAlign(void **p, int alignment, int size){ - if(*p) free(*p); - *p= memalign(alignment, size); - memset(*p, 0, size); + if(*p) free(*p); + *p= memalign(alignment, size); + memset(*p, 0, size); } static void reallocBuffers(PPContext *c, int width, int height, int stride, int qpStride){ - int mbWidth = (width+15)>>4; - int mbHeight= (height+15)>>4; - int i; + int mbWidth = (width+15)>>4; + int mbHeight= (height+15)>>4; + int i; - c->stride= stride; - c->qpStride= qpStride; + c->stride= stride; + c->qpStride= qpStride; - reallocAlign((void **)&c->tempDst, 8, stride*24); - reallocAlign((void **)&c->tempSrc, 8, stride*24); - reallocAlign((void **)&c->tempBlocks, 8, 2*16*8); - reallocAlign((void **)&c->yHistogram, 8, 256*sizeof(uint64_t)); - for(i=0; i<256; i++) - c->yHistogram[i]= width*height/64*15/256; + reallocAlign((void **)&c->tempDst, 8, stride*24); + reallocAlign((void **)&c->tempSrc, 8, stride*24); + reallocAlign((void **)&c->tempBlocks, 8, 2*16*8); + reallocAlign((void **)&c->yHistogram, 8, 256*sizeof(uint64_t)); + for(i=0; i<256; i++) + c->yHistogram[i]= width*height/64*15/256; - for(i=0; i<3; i++) - { - //Note:the +17*1024 is just there so i dont have to worry about r/w over te end - reallocAlign((void **)&c->tempBlured[i], 8, stride*mbHeight*16 + 17*1024); - reallocAlign((void **)&c->tempBluredPast[i], 8, 256*((height+7)&(~7))/2 + 17*1024);//FIXME size - } + for(i=0; i<3; i++) + { + //Note:the +17*1024 is just there so i dont have to worry about r/w over te end + reallocAlign((void **)&c->tempBlured[i], 8, stride*mbHeight*16 + 17*1024); + reallocAlign((void **)&c->tempBluredPast[i], 8, 256*((height+7)&(~7))/2 + 17*1024);//FIXME size + } - reallocAlign((void **)&c->deintTemp, 8, 2*width+32); - reallocAlign((void **)&c->nonBQPTable, 8, qpStride*mbHeight*sizeof(QP_STORE_T)); - reallocAlign((void **)&c->stdQPTable, 8, qpStride*mbHeight*sizeof(QP_STORE_T)); - reallocAlign((void **)&c->forcedQPTable, 8, mbWidth*sizeof(QP_STORE_T)); + reallocAlign((void **)&c->deintTemp, 8, 2*width+32); + reallocAlign((void **)&c->nonBQPTable, 8, qpStride*mbHeight*sizeof(QP_STORE_T)); + reallocAlign((void **)&c->stdQPTable, 8, qpStride*mbHeight*sizeof(QP_STORE_T)); + reallocAlign((void **)&c->forcedQPTable, 8, mbWidth*sizeof(QP_STORE_T)); } static void global_init(void){ - int i; - memset(clip_table, 0, 256); - for(i=256; i<512; i++) - clip_table[i]= i; - memset(clip_table+512, 0, 256); + int i; + memset(clip_table, 0, 256); + for(i=256; i<512; i++) + clip_table[i]= i; + memset(clip_table+512, 0, 256); } pp_context_t *pp_get_context(int width, int height, int cpuCaps){ - PPContext *c= memalign(32, sizeof(PPContext)); - int stride= (width+15)&(~15); //assumed / will realloc if needed - int qpStride= (width+15)/16 + 2; //assumed / will realloc if needed + PPContext *c= memalign(32, sizeof(PPContext)); + int stride= (width+15)&(~15); //assumed / will realloc if needed + int qpStride= (width+15)/16 + 2; //assumed / will realloc if needed - global_init(); + global_init(); - memset(c, 0, sizeof(PPContext)); - c->cpuCaps= cpuCaps; - if(cpuCaps&PP_FORMAT){ - c->hChromaSubSample= cpuCaps&0x3; - c->vChromaSubSample= (cpuCaps>>4)&0x3; - }else{ - c->hChromaSubSample= 1; - c->vChromaSubSample= 1; - } + memset(c, 0, sizeof(PPContext)); + c->cpuCaps= cpuCaps; + if(cpuCaps&PP_FORMAT){ + c->hChromaSubSample= cpuCaps&0x3; + c->vChromaSubSample= (cpuCaps>>4)&0x3; + }else{ + c->hChromaSubSample= 1; + c->vChromaSubSample= 1; + } - reallocBuffers(c, width, height, stride, qpStride); + reallocBuffers(c, width, height, stride, qpStride); - c->frameNum=-1; + c->frameNum=-1; - return c; + return c; } void pp_free_context(void *vc){ - PPContext *c = (PPContext*)vc; - int i; + PPContext *c = (PPContext*)vc; + int i; - for(i=0; i<3; i++) free(c->tempBlured[i]); - for(i=0; i<3; i++) free(c->tempBluredPast[i]); + for(i=0; i<3; i++) free(c->tempBlured[i]); + for(i=0; i<3; i++) free(c->tempBluredPast[i]); - free(c->tempBlocks); - free(c->yHistogram); - free(c->tempDst); - free(c->tempSrc); - free(c->deintTemp); - free(c->stdQPTable); - free(c->nonBQPTable); - free(c->forcedQPTable); + free(c->tempBlocks); + free(c->yHistogram); + free(c->tempDst); + free(c->tempSrc); + free(c->deintTemp); + free(c->stdQPTable); + free(c->nonBQPTable); + free(c->forcedQPTable); - memset(c, 0, sizeof(PPContext)); + memset(c, 0, sizeof(PPContext)); - free(c); + free(c); } void pp_postprocess(uint8_t * src[3], int srcStride[3], uint8_t * dst[3], int dstStride[3], int width, int height, QP_STORE_T *QP_store, int QPStride, - pp_mode_t *vm, void *vc, int pict_type) + pp_mode_t *vm, void *vc, int pict_type) { - int mbWidth = (width+15)>>4; - int mbHeight= (height+15)>>4; - PPMode *mode = (PPMode*)vm; - PPContext *c = (PPContext*)vc; - int minStride= MAX(ABS(srcStride[0]), ABS(dstStride[0])); - int absQPStride = ABS(QPStride); + int mbWidth = (width+15)>>4; + int mbHeight= (height+15)>>4; + PPMode *mode = (PPMode*)vm; + PPContext *c = (PPContext*)vc; + int minStride= MAX(ABS(srcStride[0]), ABS(dstStride[0])); + int absQPStride = ABS(QPStride); - // c->stride and c->QPStride are always positive - if(c->stride < minStride || c->qpStride < absQPStride) - reallocBuffers(c, width, height, - MAX(minStride, c->stride), - MAX(c->qpStride, absQPStride)); + // c->stride and c->QPStride are always positive + if(c->stride < minStride || c->qpStride < absQPStride) + reallocBuffers(c, width, height, + MAX(minStride, c->stride), + MAX(c->qpStride, absQPStride)); - if(QP_store==NULL || (mode->lumMode & FORCE_QUANT)) - { - int i; - QP_store= c->forcedQPTable; - absQPStride = QPStride = 0; - if(mode->lumMode & FORCE_QUANT) - for(i=0; i<mbWidth; i++) QP_store[i]= mode->forcedQuant; - else - for(i=0; i<mbWidth; i++) QP_store[i]= 1; - } + if(QP_store==NULL || (mode->lumMode & FORCE_QUANT)) + { + int i; + QP_store= c->forcedQPTable; + absQPStride = QPStride = 0; + if(mode->lumMode & FORCE_QUANT) + for(i=0; i<mbWidth; i++) QP_store[i]= mode->forcedQuant; + else + for(i=0; i<mbWidth; i++) QP_store[i]= 1; + } //printf("pict_type:%d\n", pict_type); - if(pict_type & PP_PICT_TYPE_QP2){ - int i; - const int count= mbHeight * absQPStride; - for(i=0; i<(count>>2); i++){ - ((uint32_t*)c->stdQPTable)[i] = (((uint32_t*)QP_store)[i]>>1) & 0x7F7F7F7F; - } - for(i<<=2; i<count; i++){ - c->stdQPTable[i] = QP_store[i]>>1; - } + if(pict_type & PP_PICT_TYPE_QP2){ + int i; + const int count= mbHeight * absQPStride; + for(i=0; i<(count>>2); i++){ + ((uint32_t*)c->stdQPTable)[i] = (((uint32_t*)QP_store)[i]>>1) & 0x7F7F7F7F; + } + for(i<<=2; i<count; i++){ + c->stdQPTable[i] = QP_store[i]>>1; + } QP_store= c->stdQPTable; - QPStride= absQPStride; - } + QPStride= absQPStride; + } if(0){ int x,y; for(y=0; y<mbHeight; y++){ - for(x=0; x<mbWidth; x++){ - printf("%2d ", QP_store[x + y*QPStride]); - } - printf("\n"); + for(x=0; x<mbWidth; x++){ + printf("%2d ", QP_store[x + y*QPStride]); + } + printf("\n"); } - printf("\n"); + printf("\n"); } - if((pict_type&7)!=3) - { - if (QPStride >= 0) { - int i; - const int count= mbHeight * QPStride; - for(i=0; i<(count>>2); i++){ - ((uint32_t*)c->nonBQPTable)[i] = ((uint32_t*)QP_store)[i] & 0x3F3F3F3F; - } - for(i<<=2; i<count; i++){ - c->nonBQPTable[i] = QP_store[i] & 0x3F; - } - } else { - int i,j; - for(i=0; i<mbHeight; i++) { - for(j=0; j<absQPStride; j++) { - c->nonBQPTable[i*absQPStride+j] = QP_store[i*QPStride+j] & 0x3F; - } - } - } - } + if((pict_type&7)!=3) + { + if (QPStride >= 0) { + int i; + const int count= mbHeight * QPStride; + for(i=0; i<(count>>2); i++){ + ((uint32_t*)c->nonBQPTable)[i] = ((uint32_t*)QP_store)[i] & 0x3F3F3F3F; + } + for(i<<=2; i<count; i++){ + c->nonBQPTable[i] = QP_store[i] & 0x3F; + } + } else { + int i,j; + for(i=0; i<mbHeight; i++) { + for(j=0; j<absQPStride; j++) { + c->nonBQPTable[i*absQPStride+j] = QP_store[i*QPStride+j] & 0x3F; + } + } + } + } - if(verbose>2) - { - printf("using npp filters 0x%X/0x%X\n", mode->lumMode, mode->chromMode); - } + if(verbose>2) + { + printf("using npp filters 0x%X/0x%X\n", mode->lumMode, mode->chromMode); + } - postProcess(src[0], srcStride[0], dst[0], dstStride[0], - width, height, QP_store, QPStride, 0, mode, c); + postProcess(src[0], srcStride[0], dst[0], dstStride[0], + width, height, QP_store, QPStride, 0, mode, c); - width = (width )>>c->hChromaSubSample; - height = (height)>>c->vChromaSubSample; + width = (width )>>c->hChromaSubSample; + height = (height)>>c->vChromaSubSample; - if(mode->chromMode) - { - postProcess(src[1], srcStride[1], dst[1], dstStride[1], - width, height, QP_store, QPStride, 1, mode, c); - postProcess(src[2], srcStride[2], dst[2], dstStride[2], - width, height, QP_store, QPStride, 2, mode, c); - } - else if(srcStride[1] == dstStride[1] && srcStride[2] == dstStride[2]) - { - linecpy(dst[1], src[1], height, srcStride[1]); - linecpy(dst[2], src[2], height, srcStride[2]); - } - else - { - int y; - for(y=0; y<height; y++) - { - memcpy(&(dst[1][y*dstStride[1]]), &(src[1][y*srcStride[1]]), width); - memcpy(&(dst[2][y*dstStride[2]]), &(src[2][y*srcStride[2]]), width); - } - } + if(mode->chromMode) + { + postProcess(src[1], srcStride[1], dst[1], dstStride[1], + width, height, QP_store, QPStride, 1, mode, c); + postProcess(src[2], srcStride[2], dst[2], dstStride[2], + width, height, QP_store, QPStride, 2, mode, c); + } + else if(srcStride[1] == dstStride[1] && srcStride[2] == dstStride[2]) + { + linecpy(dst[1], src[1], height, srcStride[1]); + linecpy(dst[2], src[2], height, srcStride[2]); + } + else + { + int y; + for(y=0; y<height; y++) + { + memcpy(&(dst[1][y*dstStride[1]]), &(src[1][y*srcStride[1]]), width); + memcpy(&(dst[2][y*dstStride[2]]), &(src[2][y*srcStride[2]]), width); + } + } }
--- a/libpostproc/postprocess.h Wed Dec 21 17:50:40 2005 +0000 +++ b/libpostproc/postprocess.h Thu Dec 22 01:10:11 2005 +0000 @@ -42,7 +42,7 @@ uint8_t * dst[3], int dstStride[3], int horizontalSize, int verticalSize, QP_STORE_T *QP_store, int QP_stride, - pp_mode_t *mode, pp_context_t *ppContext, int pict_type); + pp_mode_t *mode, pp_context_t *ppContext, int pict_type); /**
--- a/libpostproc/postprocess_altivec_template.c Wed Dec 21 17:50:40 2005 +0000 +++ b/libpostproc/postprocess_altivec_template.c Thu Dec 22 01:10:11 2005 +0000 @@ -26,35 +26,35 @@ #endif #define ALTIVEC_TRANSPOSE_8x8_SHORT(src_a,src_b,src_c,src_d,src_e,src_f,src_g,src_h) \ - do { \ - __typeof__(src_a) tempA1, tempB1, tempC1, tempD1; \ - __typeof__(src_a) tempE1, tempF1, tempG1, tempH1; \ - __typeof__(src_a) tempA2, tempB2, tempC2, tempD2; \ - __typeof__(src_a) tempE2, tempF2, tempG2, tempH2; \ - tempA1 = vec_mergeh (src_a, src_e); \ - tempB1 = vec_mergel (src_a, src_e); \ - tempC1 = vec_mergeh (src_b, src_f); \ - tempD1 = vec_mergel (src_b, src_f); \ - tempE1 = vec_mergeh (src_c, src_g); \ - tempF1 = vec_mergel (src_c, src_g); \ - tempG1 = vec_mergeh (src_d, src_h); \ - tempH1 = vec_mergel (src_d, src_h); \ - tempA2 = vec_mergeh (tempA1, tempE1); \ - tempB2 = vec_mergel (tempA1, tempE1); \ - tempC2 = vec_mergeh (tempB1, tempF1); \ - tempD2 = vec_mergel (tempB1, tempF1); \ - tempE2 = vec_mergeh (tempC1, tempG1); \ - tempF2 = vec_mergel (tempC1, tempG1); \ - tempG2 = vec_mergeh (tempD1, tempH1); \ - tempH2 = vec_mergel (tempD1, tempH1); \ - src_a = vec_mergeh (tempA2, tempE2); \ - src_b = vec_mergel (tempA2, tempE2); \ - src_c = vec_mergeh (tempB2, tempF2); \ - src_d = vec_mergel (tempB2, tempF2); \ - src_e = vec_mergeh (tempC2, tempG2); \ - src_f = vec_mergel (tempC2, tempG2); \ - src_g = vec_mergeh (tempD2, tempH2); \ - src_h = vec_mergel (tempD2, tempH2); \ + do { \ + __typeof__(src_a) tempA1, tempB1, tempC1, tempD1; \ + __typeof__(src_a) tempE1, tempF1, tempG1, tempH1; \ + __typeof__(src_a) tempA2, tempB2, tempC2, tempD2; \ + __typeof__(src_a) tempE2, tempF2, tempG2, tempH2; \ + tempA1 = vec_mergeh (src_a, src_e); \ + tempB1 = vec_mergel (src_a, src_e); \ + tempC1 = vec_mergeh (src_b, src_f); \ + tempD1 = vec_mergel (src_b, src_f); \ + tempE1 = vec_mergeh (src_c, src_g); \ + tempF1 = vec_mergel (src_c, src_g); \ + tempG1 = vec_mergeh (src_d, src_h); \ + tempH1 = vec_mergel (src_d, src_h); \ + tempA2 = vec_mergeh (tempA1, tempE1); \ + tempB2 = vec_mergel (tempA1, tempE1); \ + tempC2 = vec_mergeh (tempB1, tempF1); \ + tempD2 = vec_mergel (tempB1, tempF1); \ + tempE2 = vec_mergeh (tempC1, tempG1); \ + tempF2 = vec_mergel (tempC1, tempG1); \ + tempG2 = vec_mergeh (tempD1, tempH1); \ + tempH2 = vec_mergel (tempD1, tempH1); \ + src_a = vec_mergeh (tempA2, tempE2); \ + src_b = vec_mergel (tempA2, tempE2); \ + src_c = vec_mergeh (tempB2, tempF2); \ + src_d = vec_mergel (tempB2, tempF2); \ + src_e = vec_mergeh (tempC2, tempG2); \ + src_f = vec_mergel (tempC2, tempG2); \ + src_g = vec_mergeh (tempD2, tempH2); \ + src_h = vec_mergel (tempD2, tempH2); \ } while (0) @@ -94,25 +94,25 @@ vector signed short v_srcAss0, v_srcAss1, v_srcAss2, v_srcAss3, v_srcAss4, v_srcAss5, v_srcAss6, v_srcAss7; -#define LOAD_LINE(i) \ - register int j##i = i * stride; \ - vector unsigned char perm##i = vec_lvsl(j##i, src2); \ - const vector unsigned char v_srcA1##i = vec_ld(j##i, src2); \ - vector unsigned char v_srcA2##i; \ - if (two_vectors) \ - v_srcA2##i = vec_ld(j##i + 16, src2); \ - const vector unsigned char v_srcA##i = \ - vec_perm(v_srcA1##i, v_srcA2##i, perm##i); \ +#define LOAD_LINE(i) \ + register int j##i = i * stride; \ + vector unsigned char perm##i = vec_lvsl(j##i, src2); \ + const vector unsigned char v_srcA1##i = vec_ld(j##i, src2); \ + vector unsigned char v_srcA2##i; \ + if (two_vectors) \ + v_srcA2##i = vec_ld(j##i + 16, src2); \ + const vector unsigned char v_srcA##i = \ + vec_perm(v_srcA1##i, v_srcA2##i, perm##i); \ v_srcAss##i = \ - (vector signed short)vec_mergeh((vector signed char)zero, \ - (vector signed char)v_srcA##i) + (vector signed short)vec_mergeh((vector signed char)zero, \ + (vector signed char)v_srcA##i) #define LOAD_LINE_ALIGNED(i) \ register int j##i = i * stride; \ const vector unsigned char v_srcA##i = vec_ld(j##i, src2); \ v_srcAss##i = \ - (vector signed short)vec_mergeh((vector signed char)zero, \ - (vector signed char)v_srcA##i) + (vector signed short)vec_mergeh((vector signed char)zero, \ + (vector signed char)v_srcA##i) // special casing the aligned case is worthwhile, as all call from // the (transposed) horizontable deblocks will be aligned, i naddition @@ -139,15 +139,15 @@ #undef LOAD_LINE #undef LOAD_LINE_ALIGNED -#define ITER(i, j) \ - const vector signed short v_diff##i = \ - vec_sub(v_srcAss##i, v_srcAss##j); \ - const vector signed short v_sum##i = \ - vec_add(v_diff##i, v_dcOffset); \ - const vector signed short v_comp##i = \ - (vector signed short)vec_cmplt((vector unsigned short)v_sum##i, \ - v_dcThreshold); \ - const vector signed short v_part##i = vec_and(mask, v_comp##i); \ +#define ITER(i, j) \ + const vector signed short v_diff##i = \ + vec_sub(v_srcAss##i, v_srcAss##j); \ + const vector signed short v_sum##i = \ + vec_add(v_diff##i, v_dcOffset); \ + const vector signed short v_comp##i = \ + (vector signed short)vec_cmplt((vector unsigned short)v_sum##i, \ + v_dcThreshold); \ + const vector signed short v_part##i = vec_and(mask, v_comp##i); \ v_numEq = vec_sum4s(v_part##i, v_numEq); ITER(0, 1); @@ -167,13 +167,13 @@ if (numEq > c->ppMode.flatnessThreshold) { const vector unsigned char mmoP1 = (const vector unsigned char) - AVV(0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, - 0x00, 0x01, 0x12, 0x13, 0x08, 0x09, 0x1A, 0x1B); + AVV(0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, + 0x00, 0x01, 0x12, 0x13, 0x08, 0x09, 0x1A, 0x1B); const vector unsigned char mmoP2 = (const vector unsigned char) - AVV(0x04, 0x05, 0x16, 0x17, 0x0C, 0x0D, 0x1E, 0x1F, - 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f); + AVV(0x04, 0x05, 0x16, 0x17, 0x0C, 0x0D, 0x1E, 0x1F, + 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f); const vector unsigned char mmoP = (const vector unsigned char) - vec_lvsl(8, (unsigned char*)0); + vec_lvsl(8, (unsigned char*)0); vector signed short mmoL1 = vec_perm(v_srcAss0, v_srcAss2, mmoP1); vector signed short mmoL2 = vec_perm(v_srcAss4, v_srcAss6, mmoP2); @@ -185,9 +185,9 @@ vector unsigned short mmoSum = (vector unsigned short)vec_add(mmoDiff, v2QP); if (vec_any_gt(mmoSum, v4QP)) - return 0; + return 0; else - return 1; + return 1; } else return 2; } @@ -218,21 +218,21 @@ vector unsigned char vbT0, vbT1, vbT2, vbT3, vbT4, vbT5, vbT6, vbT7, vbT8, vbT9; #define LOAD_LINE(i) \ - const vector unsigned char perml##i = \ - vec_lvsl(i * stride, src2); \ + const vector unsigned char perml##i = \ + vec_lvsl(i * stride, src2); \ vbA##i = vec_ld(i * stride, src2); \ vbB##i = vec_ld(i * stride + 16, src2); \ vbT##i = vec_perm(vbA##i, vbB##i, perml##i); \ vb##i = \ - (vector signed short)vec_mergeh((vector unsigned char)zero, \ - (vector unsigned char)vbT##i) + (vector signed short)vec_mergeh((vector unsigned char)zero, \ + (vector unsigned char)vbT##i) #define LOAD_LINE_ALIGNED(i) \ register int j##i = i * stride; \ vbT##i = vec_ld(j##i, src2); \ vb##i = \ - (vector signed short)vec_mergeh((vector signed char)zero, \ - (vector signed char)vbT##i) + (vector signed short)vec_mergeh((vector signed char)zero, \ + (vector signed char)vbT##i) // special casing the aligned case is worthwhile, as all call from // the (transposed) horizontable deblocks will be aligned, in addition @@ -308,11 +308,11 @@ const vector signed short temp91 = vec_sub(v_sumsB8, vb5); const vector signed short v_sumsB9 = vec_add(temp91, v_last); -#define COMPUTE_VR(i, j, k) \ - const vector signed short temps1##i = \ - vec_add(v_sumsB##i, v_sumsB##k); \ - const vector signed short temps2##i = \ - vec_mladd(vb##j, (vector signed short)v_2, temps1##i); \ +#define COMPUTE_VR(i, j, k) \ + const vector signed short temps1##i = \ + vec_add(v_sumsB##i, v_sumsB##k); \ + const vector signed short temps2##i = \ + vec_mladd(vb##j, (vector signed short)v_2, temps1##i); \ const vector signed short vr##j = vec_sra(temps2##i, v_4) COMPUTE_VR(0, 1, 2); @@ -326,31 +326,31 @@ const vector signed char neg1 = vec_splat_s8(-1); const vector unsigned char permHH = (const vector unsigned char)AVV(0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, - 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); + 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); -#define PACK_AND_STORE(i) \ - const vector unsigned char perms##i = \ - vec_lvsr(i * stride, src2); \ - const vector unsigned char vf##i = \ - vec_packsu(vr##i, (vector signed short)zero); \ - const vector unsigned char vg##i = \ - vec_perm(vf##i, vbT##i, permHH); \ - const vector unsigned char mask##i = \ - vec_perm((vector unsigned char)zero, (vector unsigned char)neg1, perms##i); \ - const vector unsigned char vg2##i = \ - vec_perm(vg##i, vg##i, perms##i); \ - const vector unsigned char svA##i = \ - vec_sel(vbA##i, vg2##i, mask##i); \ - const vector unsigned char svB##i = \ - vec_sel(vg2##i, vbB##i, mask##i); \ - vec_st(svA##i, i * stride, src2); \ +#define PACK_AND_STORE(i) \ + const vector unsigned char perms##i = \ + vec_lvsr(i * stride, src2); \ + const vector unsigned char vf##i = \ + vec_packsu(vr##i, (vector signed short)zero); \ + const vector unsigned char vg##i = \ + vec_perm(vf##i, vbT##i, permHH); \ + const vector unsigned char mask##i = \ + vec_perm((vector unsigned char)zero, (vector unsigned char)neg1, perms##i); \ + const vector unsigned char vg2##i = \ + vec_perm(vg##i, vg##i, perms##i); \ + const vector unsigned char svA##i = \ + vec_sel(vbA##i, vg2##i, mask##i); \ + const vector unsigned char svB##i = \ + vec_sel(vg2##i, vbB##i, mask##i); \ + vec_st(svA##i, i * stride, src2); \ vec_st(svB##i, i * stride + 16, src2) -#define PACK_AND_STORE_ALIGNED(i) \ - const vector unsigned char vf##i = \ - vec_packsu(vr##i, (vector signed short)zero); \ - const vector unsigned char vg##i = \ - vec_perm(vf##i, vbT##i, permHH); \ +#define PACK_AND_STORE_ALIGNED(i) \ + const vector unsigned char vf##i = \ + vec_packsu(vr##i, (vector signed short)zero); \ + const vector unsigned char vg##i = \ + vec_perm(vf##i, vbT##i, permHH); \ vec_st(vg##i, i * stride, src2) // special casing the aligned case is worthwhile, as all call from @@ -398,17 +398,17 @@ vqp = vec_splat(vqp, 0); #define LOAD_LINE(i) \ - const vector unsigned char perm##i = \ - vec_lvsl(i * stride, src2); \ - const vector unsigned char vbA##i = \ - vec_ld(i * stride, src2); \ - const vector unsigned char vbB##i = \ - vec_ld(i * stride + 16, src2); \ - const vector unsigned char vbT##i = \ - vec_perm(vbA##i, vbB##i, perm##i); \ - const vector signed short vb##i = \ - (vector signed short)vec_mergeh((vector unsigned char)zero, \ - (vector unsigned char)vbT##i) + const vector unsigned char perm##i = \ + vec_lvsl(i * stride, src2); \ + const vector unsigned char vbA##i = \ + vec_ld(i * stride, src2); \ + const vector unsigned char vbB##i = \ + vec_ld(i * stride + 16, src2); \ + const vector unsigned char vbT##i = \ + vec_perm(vbA##i, vbB##i, perm##i); \ + const vector signed short vb##i = \ + (vector signed short)vec_mergeh((vector unsigned char)zero, \ + (vector unsigned char)vbT##i) src2 += stride*3; @@ -426,7 +426,7 @@ const vector signed short v_2 = vec_splat_s16(2); const vector signed short v_5 = vec_splat_s16(5); const vector signed short v_32 = vec_sl(v_1, - (vector unsigned short)v_5); + (vector unsigned short)v_5); /* middle energy */ const vector signed short l3minusl6 = vec_sub(vb3, vb6); const vector signed short l5minusl4 = vec_sub(vb5, vb4); @@ -483,22 +483,22 @@ const vector signed char neg1 = vec_splat_s8(-1); const vector unsigned char permHH = (const vector unsigned char)AVV(0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, - 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); + 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); -#define STORE(i) \ - const vector unsigned char perms##i = \ - vec_lvsr(i * stride, src2); \ - const vector unsigned char vg##i = \ - vec_perm(st##i, vbT##i, permHH); \ - const vector unsigned char mask##i = \ - vec_perm((vector unsigned char)zero, (vector unsigned char)neg1, perms##i); \ - const vector unsigned char vg2##i = \ - vec_perm(vg##i, vg##i, perms##i); \ - const vector unsigned char svA##i = \ - vec_sel(vbA##i, vg2##i, mask##i); \ - const vector unsigned char svB##i = \ - vec_sel(vg2##i, vbB##i, mask##i); \ - vec_st(svA##i, i * stride, src2); \ +#define STORE(i) \ + const vector unsigned char perms##i = \ + vec_lvsr(i * stride, src2); \ + const vector unsigned char vg##i = \ + vec_perm(st##i, vbT##i, permHH); \ + const vector unsigned char mask##i = \ + vec_perm((vector unsigned char)zero, (vector unsigned char)neg1, perms##i); \ + const vector unsigned char vg2##i = \ + vec_perm(vg##i, vg##i, perms##i); \ + const vector unsigned char svA##i = \ + vec_sel(vbA##i, vg2##i, mask##i); \ + const vector unsigned char svB##i = \ + vec_sel(vg2##i, vbB##i, mask##i); \ + vec_st(svA##i, i * stride, src2); \ vec_st(svB##i, i * stride + 16, src2) STORE(4); @@ -522,11 +522,11 @@ dt[0] = deringThreshold; v_dt = vec_splat(vec_ld(0, dt), 0); -#define LOAD_LINE(i) \ - const vector unsigned char perm##i = \ - vec_lvsl(i * stride, srcCopy); \ - vector unsigned char sA##i = vec_ld(i * stride, srcCopy); \ - vector unsigned char sB##i = vec_ld(i * stride + 16, srcCopy); \ +#define LOAD_LINE(i) \ + const vector unsigned char perm##i = \ + vec_lvsl(i * stride, srcCopy); \ + vector unsigned char sA##i = vec_ld(i * stride, srcCopy); \ + vector unsigned char sB##i = vec_ld(i * stride + 16, srcCopy); \ vector unsigned char src##i = vec_perm(sA##i, sB##i, perm##i) LOAD_LINE(0); @@ -545,13 +545,13 @@ { const vector unsigned char trunc_perm = (vector unsigned char) AVV(0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, - 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18); + 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18); const vector unsigned char trunc_src12 = vec_perm(src1, src2, trunc_perm); const vector unsigned char trunc_src34 = vec_perm(src3, src4, trunc_perm); const vector unsigned char trunc_src56 = vec_perm(src5, src6, trunc_perm); const vector unsigned char trunc_src78 = vec_perm(src7, src8, trunc_perm); -#define EXTRACT(op) do { \ +#define EXTRACT(op) do { \ const vector unsigned char s##op##_1 = vec_##op(trunc_src12, trunc_src34); \ const vector unsigned char s##op##_2 = vec_##op(trunc_src56, trunc_src78); \ const vector unsigned char s##op##_6 = vec_##op(s##op##_1, s##op##_2); \ @@ -584,29 +584,29 @@ { const vector unsigned short mask1 = (vector unsigned short) AVV(0x0001, 0x0002, 0x0004, 0x0008, - 0x0010, 0x0020, 0x0040, 0x0080); + 0x0010, 0x0020, 0x0040, 0x0080); const vector unsigned short mask2 = (vector unsigned short) AVV(0x0100, 0x0200, 0x0000, 0x0000, - 0x0000, 0x0000, 0x0000, 0x0000); + 0x0000, 0x0000, 0x0000, 0x0000); const vector unsigned int vuint32_16 = vec_sl(vec_splat_u32(1), vec_splat_u32(4)); const vector unsigned int vuint32_1 = vec_splat_u32(1); -#define COMPARE(i) \ - vector signed int sum##i; \ - do { \ - const vector unsigned char cmp##i = \ - (vector unsigned char)vec_cmpgt(src##i, v_avg); \ - const vector unsigned short cmpHi##i = \ - (vector unsigned short)vec_mergeh(cmp##i, cmp##i); \ - const vector unsigned short cmpLi##i = \ - (vector unsigned short)vec_mergel(cmp##i, cmp##i); \ - const vector signed short cmpHf##i = \ - (vector signed short)vec_and(cmpHi##i, mask1); \ - const vector signed short cmpLf##i = \ - (vector signed short)vec_and(cmpLi##i, mask2); \ - const vector signed int sump##i = vec_sum4s(cmpHf##i, zero); \ - const vector signed int sumq##i = vec_sum4s(cmpLf##i, sump##i); \ +#define COMPARE(i) \ + vector signed int sum##i; \ + do { \ + const vector unsigned char cmp##i = \ + (vector unsigned char)vec_cmpgt(src##i, v_avg); \ + const vector unsigned short cmpHi##i = \ + (vector unsigned short)vec_mergeh(cmp##i, cmp##i); \ + const vector unsigned short cmpLi##i = \ + (vector unsigned short)vec_mergel(cmp##i, cmp##i); \ + const vector signed short cmpHf##i = \ + (vector signed short)vec_and(cmpHi##i, mask1); \ + const vector signed short cmpLf##i = \ + (vector signed short)vec_and(cmpLi##i, mask2); \ + const vector signed int sump##i = vec_sum4s(cmpHf##i, zero); \ + const vector signed int sumq##i = vec_sum4s(cmpLf##i, sump##i); \ sum##i = vec_sums(sumq##i, zero); } while (0) COMPARE(0); @@ -643,11 +643,11 @@ const vector signed int t2B = vec_or(sumB, tB); const vector signed int t2C = vec_or(sumC, tC); const vector signed int t3A = vec_and(vec_sra(t2A, vuint32_1), - vec_sl(t2A, vuint32_1)); + vec_sl(t2A, vuint32_1)); const vector signed int t3B = vec_and(vec_sra(t2B, vuint32_1), - vec_sl(t2B, vuint32_1)); + vec_sl(t2B, vuint32_1)); const vector signed int t3C = vec_and(vec_sra(t2C, vuint32_1), - vec_sl(t2C, vuint32_1)); + vec_sl(t2C, vuint32_1)); const vector signed int yA = vec_and(t2A, t3A); const vector signed int yB = vec_and(t2B, t3B); const vector signed int yC = vec_and(t2C, t3C); @@ -659,15 +659,15 @@ const vector signed int sumBd4 = vec_perm(yB, yC, strangeperm1); const vector signed int sumBd8 = vec_perm(yB, yC, strangeperm2); const vector signed int sumAp = vec_and(yA, - vec_and(sumAd4,sumAd8)); + vec_and(sumAd4,sumAd8)); const vector signed int sumBp = vec_and(yB, - vec_and(sumBd4,sumBd8)); + vec_and(sumBd4,sumBd8)); sumA2 = vec_or(sumAp, - vec_sra(sumAp, - vuint32_16)); + vec_sra(sumAp, + vuint32_16)); sumB2 = vec_or(sumBp, - vec_sra(sumBp, - vuint32_16)); + vec_sra(sumBp, + vuint32_16)); } vec_st(sumA2, 0, S); vec_st(sumB2, 16, S); @@ -686,84 +686,84 @@ const vector unsigned char permA1 = (vector unsigned char) AVV(0x00, 0x01, 0x02, 0x10, 0x11, 0x12, 0x1F, 0x1F, - 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F); + 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F); const vector unsigned char permA2 = (vector unsigned char) AVV(0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x10, 0x11, - 0x12, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F); + 0x12, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F, 0x1F); const vector unsigned char permA1inc = (vector unsigned char) AVV(0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x00, 0x00, - 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); const vector unsigned char permA2inc = (vector unsigned char) AVV(0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x01, - 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); + 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); const vector unsigned char magic = (vector unsigned char) AVV(0x01, 0x02, 0x01, 0x02, 0x04, 0x02, 0x01, 0x02, - 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); + 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); const vector unsigned char extractPerm = (vector unsigned char) AVV(0x10, 0x10, 0x10, 0x01, 0x10, 0x10, 0x10, 0x01, - 0x10, 0x10, 0x10, 0x01, 0x10, 0x10, 0x10, 0x01); + 0x10, 0x10, 0x10, 0x01, 0x10, 0x10, 0x10, 0x01); const vector unsigned char extractPermInc = (vector unsigned char) AVV(0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, - 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01); + 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01); const vector unsigned char identity = vec_lvsl(0,(unsigned char *)0); const vector unsigned char tenRight = (vector unsigned char) AVV(0x00, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, - 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); const vector unsigned char eightLeft = (vector unsigned char) AVV(0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, - 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08); + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08); -#define F_INIT(i) \ - vector unsigned char tenRightM##i = tenRight; \ - vector unsigned char permA1M##i = permA1; \ - vector unsigned char permA2M##i = permA2; \ +#define F_INIT(i) \ + vector unsigned char tenRightM##i = tenRight; \ + vector unsigned char permA1M##i = permA1; \ + vector unsigned char permA2M##i = permA2; \ vector unsigned char extractPermM##i = extractPerm -#define F2(i, j, k, l) \ - if (S[i] & (1 << (l+1))) { \ - const vector unsigned char a_##j##_A##l = \ - vec_perm(src##i, src##j, permA1M##i); \ - const vector unsigned char a_##j##_B##l = \ - vec_perm(a_##j##_A##l, src##k, permA2M##i); \ - const vector signed int a_##j##_sump##l = \ - (vector signed int)vec_msum(a_##j##_B##l, magic, \ - (vector unsigned int)zero); \ - vector signed int F_##j##_##l = \ - vec_sr(vec_sums(a_##j##_sump##l, vsint32_8), vuint32_4); \ - F_##j##_##l = vec_splat(F_##j##_##l, 3); \ - const vector signed int p_##j##_##l = \ - (vector signed int)vec_perm(src##j, \ - (vector unsigned char)zero, \ - extractPermM##i); \ - const vector signed int sum_##j##_##l = vec_add( p_##j##_##l, vQP2); \ - const vector signed int diff_##j##_##l = vec_sub( p_##j##_##l, vQP2); \ - vector signed int newpm_##j##_##l; \ - if (vec_all_lt(sum_##j##_##l, F_##j##_##l)) \ - newpm_##j##_##l = sum_##j##_##l; \ - else if (vec_all_gt(diff_##j##_##l, F_##j##_##l)) \ - newpm_##j##_##l = diff_##j##_##l; \ - else newpm_##j##_##l = F_##j##_##l; \ - const vector unsigned char newpm2_##j##_##l = \ - vec_splat((vector unsigned char)newpm_##j##_##l, 15); \ - const vector unsigned char mask##j##l = vec_add(identity, \ - tenRightM##i); \ - src##j = vec_perm(src##j, newpm2_##j##_##l, mask##j##l); \ - } \ - permA1M##i = vec_add(permA1M##i, permA1inc); \ - permA2M##i = vec_add(permA2M##i, permA2inc); \ - tenRightM##i = vec_sro(tenRightM##i, eightLeft); \ +#define F2(i, j, k, l) \ + if (S[i] & (1 << (l+1))) { \ + const vector unsigned char a_##j##_A##l = \ + vec_perm(src##i, src##j, permA1M##i); \ + const vector unsigned char a_##j##_B##l = \ + vec_perm(a_##j##_A##l, src##k, permA2M##i); \ + const vector signed int a_##j##_sump##l = \ + (vector signed int)vec_msum(a_##j##_B##l, magic, \ + (vector unsigned int)zero); \ + vector signed int F_##j##_##l = \ + vec_sr(vec_sums(a_##j##_sump##l, vsint32_8), vuint32_4); \ + F_##j##_##l = vec_splat(F_##j##_##l, 3); \ + const vector signed int p_##j##_##l = \ + (vector signed int)vec_perm(src##j, \ + (vector unsigned char)zero, \ + extractPermM##i); \ + const vector signed int sum_##j##_##l = vec_add( p_##j##_##l, vQP2);\ + const vector signed int diff_##j##_##l = vec_sub( p_##j##_##l, vQP2);\ + vector signed int newpm_##j##_##l; \ + if (vec_all_lt(sum_##j##_##l, F_##j##_##l)) \ + newpm_##j##_##l = sum_##j##_##l; \ + else if (vec_all_gt(diff_##j##_##l, F_##j##_##l)) \ + newpm_##j##_##l = diff_##j##_##l; \ + else newpm_##j##_##l = F_##j##_##l; \ + const vector unsigned char newpm2_##j##_##l = \ + vec_splat((vector unsigned char)newpm_##j##_##l, 15); \ + const vector unsigned char mask##j##l = vec_add(identity, \ + tenRightM##i); \ + src##j = vec_perm(src##j, newpm2_##j##_##l, mask##j##l); \ + } \ + permA1M##i = vec_add(permA1M##i, permA1inc); \ + permA2M##i = vec_add(permA2M##i, permA2inc); \ + tenRightM##i = vec_sro(tenRightM##i, eightLeft); \ extractPermM##i = vec_add(extractPermM##i, extractPermInc) -#define ITER(i, j, k) \ - F_INIT(i); \ - F2(i, j, k, 0); \ - F2(i, j, k, 1); \ - F2(i, j, k, 2); \ - F2(i, j, k, 3); \ - F2(i, j, k, 4); \ - F2(i, j, k, 5); \ - F2(i, j, k, 6); \ +#define ITER(i, j, k) \ + F_INIT(i); \ + F2(i, j, k, 0); \ + F2(i, j, k, 1); \ + F2(i, j, k, 2); \ + F2(i, j, k, 3); \ + F2(i, j, k, 4); \ + F2(i, j, k, 5); \ + F2(i, j, k, 6); \ F2(i, j, k, 7) ITER(0, 1, 2); @@ -777,16 +777,16 @@ const vector signed char neg1 = vec_splat_s8(-1); -#define STORE_LINE(i) \ - const vector unsigned char permST##i = \ - vec_lvsr(i * stride, srcCopy); \ - const vector unsigned char maskST##i = \ - vec_perm((vector unsigned char)zero, \ - (vector unsigned char)neg1, permST##i); \ - src##i = vec_perm(src##i ,src##i, permST##i); \ - sA##i= vec_sel(sA##i, src##i, maskST##i); \ - sB##i= vec_sel(src##i, sB##i, maskST##i); \ - vec_st(sA##i, i * stride, srcCopy); \ +#define STORE_LINE(i) \ + const vector unsigned char permST##i = \ + vec_lvsr(i * stride, srcCopy); \ + const vector unsigned char maskST##i = \ + vec_perm((vector unsigned char)zero, \ + (vector unsigned char)neg1, permST##i); \ + src##i = vec_perm(src##i ,src##i, permST##i); \ + sA##i= vec_sel(sA##i, src##i, maskST##i); \ + sB##i= vec_sel(src##i, sB##i, maskST##i); \ + vec_st(sA##i, i * stride, srcCopy); \ vec_st(sB##i, i * stride + 16, srcCopy) STORE_LINE(1); @@ -808,7 +808,7 @@ #define do_a_deblock_altivec(a...) do_a_deblock_C(a) static inline void RENAME(tempNoiseReducer)(uint8_t *src, int stride, - uint8_t *tempBlured, uint32_t *tempBluredPast, int *maxNoise) + uint8_t *tempBlured, uint32_t *tempBluredPast, int *maxNoise) { const vector signed int zero = vec_splat_s32(0); const vector signed short vsint16_1 = vec_splat_s16(1); @@ -820,16 +820,16 @@ tempBluredPast[128]= maxNoise[1]; tempBluredPast[129]= maxNoise[2]; -#define LOAD_LINE(src, i) \ - register int j##src##i = i * stride; \ - vector unsigned char perm##src##i = vec_lvsl(j##src##i, src); \ - const vector unsigned char v_##src##A1##i = vec_ld(j##src##i, src); \ +#define LOAD_LINE(src, i) \ + register int j##src##i = i * stride; \ + vector unsigned char perm##src##i = vec_lvsl(j##src##i, src); \ + const vector unsigned char v_##src##A1##i = vec_ld(j##src##i, src); \ const vector unsigned char v_##src##A2##i = vec_ld(j##src##i + 16, src); \ - const vector unsigned char v_##src##A##i = \ - vec_perm(v_##src##A1##i, v_##src##A2##i, perm##src##i); \ - vector signed short v_##src##Ass##i = \ - (vector signed short)vec_mergeh((vector signed char)zero, \ - (vector signed char)v_##src##A##i) + const vector unsigned char v_##src##A##i = \ + vec_perm(v_##src##A1##i, v_##src##A2##i, perm##src##i); \ + vector signed short v_##src##Ass##i = \ + (vector signed short)vec_mergeh((vector signed char)zero, \ + (vector signed char)v_##src##A##i) LOAD_LINE(src, 0); LOAD_LINE(src, 1); @@ -850,10 +850,10 @@ LOAD_LINE(tempBlured, 7); #undef LOAD_LINE -#define ACCUMULATE_DIFFS(i) \ - vector signed short v_d##i = vec_sub(v_tempBluredAss##i, \ - v_srcAss##i); \ - v_dp = vec_msums(v_d##i, v_d##i, v_dp); \ +#define ACCUMULATE_DIFFS(i) \ + vector signed short v_d##i = vec_sub(v_tempBluredAss##i, \ + v_srcAss##i); \ + v_dp = vec_msums(v_d##i, v_d##i, v_dp); \ v_sysdp = vec_msums(v_d##i, vsint16_1, v_sysdp) ACCUMULATE_DIFFS(0); @@ -916,12 +916,12 @@ const vector signed short vsint16_4 = vec_splat_s16(4); const vector unsigned short vuint16_3 = vec_splat_u16(3); -#define OP(i) \ - const vector signed short v_temp##i = \ - vec_mladd(v_tempBluredAss##i, \ - vsint16_7, v_srcAss##i); \ - const vector signed short v_temp2##i = \ - vec_add(v_temp##i, vsint16_4); \ +#define OP(i) \ + const vector signed short v_temp##i = \ + vec_mladd(v_tempBluredAss##i, \ + vsint16_7, v_srcAss##i); \ + const vector signed short v_temp2##i = \ + vec_add(v_temp##i, vsint16_4); \ v_tempBluredAss##i = vec_sr(v_temp2##i, vuint16_3) OP(0); @@ -937,12 +937,12 @@ const vector signed short vsint16_3 = vec_splat_s16(3); const vector signed short vsint16_2 = vec_splat_s16(2); -#define OP(i) \ - const vector signed short v_temp##i = \ - vec_mladd(v_tempBluredAss##i, \ - vsint16_3, v_srcAss##i); \ - const vector signed short v_temp2##i = \ - vec_add(v_temp##i, vsint16_2); \ +#define OP(i) \ + const vector signed short v_temp##i = \ + vec_mladd(v_tempBluredAss##i, \ + vsint16_3, v_srcAss##i); \ + const vector signed short v_temp2##i = \ + vec_add(v_temp##i, vsint16_2); \ v_tempBluredAss##i = vec_sr(v_temp2##i, (vector unsigned short)vsint16_2) OP(0); @@ -959,24 +959,24 @@ const vector signed char neg1 = vec_splat_s8(-1); const vector unsigned char permHH = (const vector unsigned char)AVV(0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, - 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); + 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); -#define PACK_AND_STORE(src, i) \ - const vector unsigned char perms##src##i = \ - vec_lvsr(i * stride, src); \ - const vector unsigned char vf##src##i = \ - vec_packsu(v_tempBluredAss##i, (vector signed short)zero); \ - const vector unsigned char vg##src##i = \ - vec_perm(vf##src##i, v_##src##A##i, permHH); \ - const vector unsigned char mask##src##i = \ +#define PACK_AND_STORE(src, i) \ + const vector unsigned char perms##src##i = \ + vec_lvsr(i * stride, src); \ + const vector unsigned char vf##src##i = \ + vec_packsu(v_tempBluredAss##i, (vector signed short)zero); \ + const vector unsigned char vg##src##i = \ + vec_perm(vf##src##i, v_##src##A##i, permHH); \ + const vector unsigned char mask##src##i = \ vec_perm((vector unsigned char)zero, (vector unsigned char)neg1, perms##src##i); \ - const vector unsigned char vg2##src##i = \ - vec_perm(vg##src##i, vg##src##i, perms##src##i); \ - const vector unsigned char svA##src##i = \ - vec_sel(v_##src##A1##i, vg2##src##i, mask##src##i); \ - const vector unsigned char svB##src##i = \ - vec_sel(vg2##src##i, v_##src##A2##i, mask##src##i); \ - vec_st(svA##src##i, i * stride, src); \ + const vector unsigned char vg2##src##i = \ + vec_perm(vg##src##i, vg##src##i, perms##src##i); \ + const vector unsigned char svA##src##i = \ + vec_sel(v_##src##A1##i, vg2##src##i, mask##src##i); \ + const vector unsigned char svB##src##i = \ + vec_sel(vg2##src##i, v_##src##A2##i, mask##src##i); \ + vec_st(svA##src##i, i * stride, src); \ vec_st(svB##src##i, i * stride + 16, src) PACK_AND_STORE(src, 0); @@ -1001,14 +1001,14 @@ static inline void transpose_16x8_char_toPackedAlign_altivec(unsigned char* dst, unsigned char* src, int stride) { const vector unsigned char zero = vec_splat_u8(0); -#define LOAD_DOUBLE_LINE(i, j) \ - vector unsigned char perm1##i = vec_lvsl(i * stride, src); \ - vector unsigned char perm2##i = vec_lvsl(j * stride, src); \ - vector unsigned char srcA##i = vec_ld(i * stride, src); \ +#define LOAD_DOUBLE_LINE(i, j) \ + vector unsigned char perm1##i = vec_lvsl(i * stride, src); \ + vector unsigned char perm2##i = vec_lvsl(j * stride, src); \ + vector unsigned char srcA##i = vec_ld(i * stride, src); \ vector unsigned char srcB##i = vec_ld(i * stride + 16, src); \ - vector unsigned char srcC##i = vec_ld(j * stride, src); \ + vector unsigned char srcC##i = vec_ld(j * stride, src); \ vector unsigned char srcD##i = vec_ld(j * stride+ 16, src); \ - vector unsigned char src##i = vec_perm(srcA##i, srcB##i, perm1##i); \ + vector unsigned char src##i = vec_perm(srcA##i, srcB##i, perm1##i); \ vector unsigned char src##j = vec_perm(srcC##i, srcD##i, perm2##i) LOAD_DOUBLE_LINE(0, 1); @@ -1107,10 +1107,10 @@ const vector unsigned char zero = vec_splat_u8(0); const vector unsigned char magic_perm = (const vector unsigned char) AVV(0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, - 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); + 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F); -#define LOAD_DOUBLE_LINE(i, j) \ - vector unsigned char src##i = vec_ld(i * 16, src); \ +#define LOAD_DOUBLE_LINE(i, j) \ + vector unsigned char src##i = vec_ld(i * 16, src); \ vector unsigned char src##j = vec_ld(j * 16, src) LOAD_DOUBLE_LINE(0, 1); @@ -1169,24 +1169,24 @@ const vector signed char neg1 = vec_splat_s8(-1); -#define STORE_DOUBLE_LINE(i, j) \ - vector unsigned char dstA##i = vec_ld(i * stride, dst); \ - vector unsigned char dstB##i = vec_ld(i * stride + 16, dst); \ - vector unsigned char dstA##j = vec_ld(j * stride, dst); \ - vector unsigned char dstB##j = vec_ld(j * stride+ 16, dst); \ - vector unsigned char align##i = vec_lvsr(i * stride, dst); \ - vector unsigned char align##j = vec_lvsr(j * stride, dst); \ - vector unsigned char mask##i = vec_perm(zero, (vector unsigned char)neg1, align##i); \ - vector unsigned char mask##j = vec_perm(zero, (vector unsigned char)neg1, align##j); \ - vector unsigned char dstR##i = vec_perm(temp##i, temp##i, align##i); \ - vector unsigned char dstR##j = vec_perm(temp##j, temp##j, align##j); \ - vector unsigned char dstAF##i = vec_sel(dstA##i, dstR##i, mask##i); \ - vector unsigned char dstBF##i = vec_sel(dstR##i, dstB##i, mask##i); \ - vector unsigned char dstAF##j = vec_sel(dstA##j, dstR##j, mask##j); \ - vector unsigned char dstBF##j = vec_sel(dstR##j, dstB##j, mask##j); \ - vec_st(dstAF##i, i * stride, dst); \ - vec_st(dstBF##i, i * stride + 16, dst); \ - vec_st(dstAF##j, j * stride, dst); \ +#define STORE_DOUBLE_LINE(i, j) \ + vector unsigned char dstA##i = vec_ld(i * stride, dst); \ + vector unsigned char dstB##i = vec_ld(i * stride + 16, dst); \ + vector unsigned char dstA##j = vec_ld(j * stride, dst); \ + vector unsigned char dstB##j = vec_ld(j * stride+ 16, dst); \ + vector unsigned char align##i = vec_lvsr(i * stride, dst); \ + vector unsigned char align##j = vec_lvsr(j * stride, dst); \ + vector unsigned char mask##i = vec_perm(zero, (vector unsigned char)neg1, align##i); \ + vector unsigned char mask##j = vec_perm(zero, (vector unsigned char)neg1, align##j); \ + vector unsigned char dstR##i = vec_perm(temp##i, temp##i, align##i); \ + vector unsigned char dstR##j = vec_perm(temp##j, temp##j, align##j); \ + vector unsigned char dstAF##i = vec_sel(dstA##i, dstR##i, mask##i); \ + vector unsigned char dstBF##i = vec_sel(dstR##i, dstB##i, mask##i); \ + vector unsigned char dstAF##j = vec_sel(dstA##j, dstR##j, mask##j); \ + vector unsigned char dstBF##j = vec_sel(dstR##j, dstB##j, mask##j); \ + vec_st(dstAF##i, i * stride, dst); \ + vec_st(dstBF##i, i * stride + 16, dst); \ + vec_st(dstAF##j, j * stride, dst); \ vec_st(dstBF##j, j * stride + 16, dst) STORE_DOUBLE_LINE(0,1);
--- a/libpostproc/postprocess_internal.h Wed Dec 21 17:50:40 2005 +0000 +++ b/libpostproc/postprocess_internal.h Thu Dec 22 01:10:11 2005 +0000 @@ -21,42 +21,42 @@ * internal api header. */ -#define V_DEBLOCK 0x01 -#define H_DEBLOCK 0x02 -#define DERING 0x04 -#define LEVEL_FIX 0x08 ///< Brightness & Contrast +#define V_DEBLOCK 0x01 +#define H_DEBLOCK 0x02 +#define DERING 0x04 +#define LEVEL_FIX 0x08 ///< Brightness & Contrast -#define LUM_V_DEBLOCK V_DEBLOCK // 1 -#define LUM_H_DEBLOCK H_DEBLOCK // 2 -#define CHROM_V_DEBLOCK (V_DEBLOCK<<4) // 16 -#define CHROM_H_DEBLOCK (H_DEBLOCK<<4) // 32 -#define LUM_DERING DERING // 4 -#define CHROM_DERING (DERING<<4) // 64 -#define LUM_LEVEL_FIX LEVEL_FIX // 8 -#define CHROM_LEVEL_FIX (LEVEL_FIX<<4) // 128 (not implemented yet) +#define LUM_V_DEBLOCK V_DEBLOCK // 1 +#define LUM_H_DEBLOCK H_DEBLOCK // 2 +#define CHROM_V_DEBLOCK (V_DEBLOCK<<4) // 16 +#define CHROM_H_DEBLOCK (H_DEBLOCK<<4) // 32 +#define LUM_DERING DERING // 4 +#define CHROM_DERING (DERING<<4) // 64 +#define LUM_LEVEL_FIX LEVEL_FIX // 8 +#define CHROM_LEVEL_FIX (LEVEL_FIX<<4) // 128 (not implemented yet) // Experimental vertical filters -#define V_X1_FILTER 0x0200 // 512 -#define V_A_DEBLOCK 0x0400 +#define V_X1_FILTER 0x0200 // 512 +#define V_A_DEBLOCK 0x0400 // Experimental horizontal filters -#define H_X1_FILTER 0x2000 // 8192 -#define H_A_DEBLOCK 0x4000 +#define H_X1_FILTER 0x2000 // 8192 +#define H_A_DEBLOCK 0x4000 /// select between full y range (255-0) or standart one (234-16) -#define FULL_Y_RANGE 0x8000 // 32768 +#define FULL_Y_RANGE 0x8000 // 32768 //Deinterlacing Filters -#define LINEAR_IPOL_DEINT_FILTER 0x10000 // 65536 -#define LINEAR_BLEND_DEINT_FILTER 0x20000 // 131072 -#define CUBIC_BLEND_DEINT_FILTER 0x8000 // (not implemented yet) -#define CUBIC_IPOL_DEINT_FILTER 0x40000 // 262144 -#define MEDIAN_DEINT_FILTER 0x80000 // 524288 -#define FFMPEG_DEINT_FILTER 0x400000 -#define LOWPASS5_DEINT_FILTER 0x800000 +#define LINEAR_IPOL_DEINT_FILTER 0x10000 // 65536 +#define LINEAR_BLEND_DEINT_FILTER 0x20000 // 131072 +#define CUBIC_BLEND_DEINT_FILTER 0x8000 // (not implemented yet) +#define CUBIC_IPOL_DEINT_FILTER 0x40000 // 262144 +#define MEDIAN_DEINT_FILTER 0x80000 // 524288 +#define FFMPEG_DEINT_FILTER 0x400000 +#define LOWPASS5_DEINT_FILTER 0x800000 -#define TEMP_NOISE_FILTER 0x100000 -#define FORCE_QUANT 0x200000 +#define TEMP_NOISE_FILTER 0x100000 +#define FORCE_QUANT 0x200000 //use if u want a faster postprocessing code //cant differentiate between chroma & luma filters (both on or both off) @@ -66,8 +66,8 @@ #if 1 static inline int CLIP(int a){ - if(a&256) return ((a)>>31)^(-1); - else return a; + if(a&256) return ((a)>>31)^(-1); + else return a; } //#define CLIP(a) (((a)&256) ? ((a)>>31)^(-1) : (a)) #elif 0 @@ -79,92 +79,92 @@ * Postprocessng filter. */ struct PPFilter{ - char *shortName; - char *longName; - int chromDefault; ///< is chrominance filtering on by default if this filter is manually activated - int minLumQuality; ///< minimum quality to turn luminance filtering on - int minChromQuality; ///< minimum quality to turn chrominance filtering on - int mask; ///< Bitmask to turn this filter on + char *shortName; + char *longName; + int chromDefault; ///< is chrominance filtering on by default if this filter is manually activated + int minLumQuality; ///< minimum quality to turn luminance filtering on + int minChromQuality; ///< minimum quality to turn chrominance filtering on + int mask; ///< Bitmask to turn this filter on }; /** * Postprocessng mode. */ typedef struct PPMode{ - int lumMode; ///< acivates filters for luminance - int chromMode; ///< acivates filters for chrominance - int error; ///< non zero on error + int lumMode; ///< acivates filters for luminance + int chromMode; ///< acivates filters for chrominance + int error; ///< non zero on error - int minAllowedY; ///< for brigtness correction - int maxAllowedY; ///< for brihtness correction - float maxClippedThreshold; ///< amount of "black" u r willing to loose to get a brightness corrected picture + int minAllowedY; ///< for brigtness correction + int maxAllowedY; ///< for brihtness correction + float maxClippedThreshold; ///< amount of "black" u r willing to loose to get a brightness corrected picture - int maxTmpNoise[3]; ///< for Temporal Noise Reducing filter (Maximal sum of abs differences) + int maxTmpNoise[3]; ///< for Temporal Noise Reducing filter (Maximal sum of abs differences) - int baseDcDiff; - int flatnessThreshold; + int baseDcDiff; + int flatnessThreshold; - int forcedQuant; ///< quantizer if FORCE_QUANT is used + int forcedQuant; ///< quantizer if FORCE_QUANT is used } PPMode; /** * postprocess context. */ typedef struct PPContext{ - uint8_t *tempBlocks; ///<used for the horizontal code + uint8_t *tempBlocks; ///<used for the horizontal code - /** - * luma histogram. - * we need 64bit here otherwise we'll going to have a problem - * after watching a black picture for 5 hours - */ - uint64_t *yHistogram; + /** + * luma histogram. + * we need 64bit here otherwise we'll going to have a problem + * after watching a black picture for 5 hours + */ + uint64_t *yHistogram; - uint64_t __attribute__((aligned(8))) packedYOffset; - uint64_t __attribute__((aligned(8))) packedYScale; + uint64_t __attribute__((aligned(8))) packedYOffset; + uint64_t __attribute__((aligned(8))) packedYScale; - /** Temporal noise reducing buffers */ - uint8_t *tempBlured[3]; - int32_t *tempBluredPast[3]; + /** Temporal noise reducing buffers */ + uint8_t *tempBlured[3]; + int32_t *tempBluredPast[3]; - /** Temporary buffers for handling the last row(s) */ - uint8_t *tempDst; - uint8_t *tempSrc; + /** Temporary buffers for handling the last row(s) */ + uint8_t *tempDst; + uint8_t *tempSrc; - uint8_t *deintTemp; + uint8_t *deintTemp; - uint64_t __attribute__((aligned(8))) pQPb; - uint64_t __attribute__((aligned(8))) pQPb2; + uint64_t __attribute__((aligned(8))) pQPb; + uint64_t __attribute__((aligned(8))) pQPb2; - uint64_t __attribute__((aligned(8))) mmxDcOffset[64]; - uint64_t __attribute__((aligned(8))) mmxDcThreshold[64]; + uint64_t __attribute__((aligned(8))) mmxDcOffset[64]; + uint64_t __attribute__((aligned(8))) mmxDcThreshold[64]; - QP_STORE_T *stdQPTable; ///< used to fix MPEG2 style qscale - QP_STORE_T *nonBQPTable; - QP_STORE_T *forcedQPTable; + QP_STORE_T *stdQPTable; ///< used to fix MPEG2 style qscale + QP_STORE_T *nonBQPTable; + QP_STORE_T *forcedQPTable; - int QP; - int nonBQP; + int QP; + int nonBQP; - int frameNum; + int frameNum; - int cpuCaps; + int cpuCaps; - int qpStride; ///<size of qp buffers (needed to realloc them if needed) - int stride; ///<size of some buffers (needed to realloc them if needed) + int qpStride; ///<size of qp buffers (needed to realloc them if needed) + int stride; ///<size of some buffers (needed to realloc them if needed) - int hChromaSubSample; - int vChromaSubSample; + int hChromaSubSample; + int vChromaSubSample; - PPMode ppMode; + PPMode ppMode; } PPContext; static inline void linecpy(void *dest, void *src, int lines, int stride) { - if (stride > 0) { - memcpy(dest, src, lines*stride); - } else { - memcpy(dest+(lines-1)*stride, src+(lines-1)*stride, -lines*stride); - } + if (stride > 0) { + memcpy(dest, src, lines*stride); + } else { + memcpy(dest+(lines-1)*stride, src+(lines-1)*stride, -lines*stride); + } }
--- a/libpostproc/postprocess_template.c Wed Dec 21 17:50:40 2005 +0000 +++ b/libpostproc/postprocess_template.c Thu Dec 22 01:10:11 2005 +0000 @@ -58,17 +58,17 @@ #define PMINUB(a,b,t) "pminub " #a ", " #b " \n\t" #elif defined (HAVE_MMX) #define PMINUB(b,a,t) \ - "movq " #a ", " #t " \n\t"\ - "psubusb " #b ", " #t " \n\t"\ - "psubb " #t ", " #a " \n\t" + "movq " #a ", " #t " \n\t"\ + "psubusb " #b ", " #t " \n\t"\ + "psubb " #t ", " #a " \n\t" #endif #ifdef HAVE_MMX2 #define PMAXUB(a,b) "pmaxub " #a ", " #b " \n\t" #elif defined (HAVE_MMX) #define PMAXUB(a,b) \ - "psubusb " #a ", " #b " \n\t"\ - "paddb " #a ", " #b " \n\t" + "psubusb " #a ", " #b " \n\t"\ + "paddb " #a ", " #b " \n\t" #endif //FIXME? |255-0| = 1 (shouldnt be a problem ...) @@ -77,109 +77,109 @@ * Check if the middle 8x8 Block in the given 8x16 block is flat */ static inline int RENAME(vertClassify)(uint8_t src[], int stride, PPContext *c){ - int numEq= 0, dcOk; - src+= stride*4; // src points to begin of the 8x8 Block + int numEq= 0, dcOk; + src+= stride*4; // src points to begin of the 8x8 Block asm volatile( - "movq %0, %%mm7 \n\t" - "movq %1, %%mm6 \n\t" + "movq %0, %%mm7 \n\t" + "movq %1, %%mm6 \n\t" : : "m" (c->mmxDcOffset[c->nonBQP]), "m" (c->mmxDcThreshold[c->nonBQP]) ); asm volatile( - "lea (%2, %3), %%"REG_a" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %1 eax eax+%2 eax+2%2 %1+4%2 ecx ecx+%2 ecx+2%2 %1+8%2 ecx+4%2 - - "movq (%2), %%mm0 \n\t" - "movq (%%"REG_a"), %%mm1 \n\t" - "movq %%mm0, %%mm3 \n\t" - "movq %%mm0, %%mm4 \n\t" + "lea (%2, %3), %%"REG_a" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %1 eax eax+%2 eax+2%2 %1+4%2 ecx ecx+%2 ecx+2%2 %1+8%2 ecx+4%2 + + "movq (%2), %%mm0 \n\t" + "movq (%%"REG_a"), %%mm1 \n\t" + "movq %%mm0, %%mm3 \n\t" + "movq %%mm0, %%mm4 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm0 \n\t" // mm0 = differnece - "paddb %%mm7, %%mm0 \n\t" - "pcmpgtb %%mm6, %%mm0 \n\t" - - "movq (%%"REG_a",%3), %%mm2 \n\t" + "psubb %%mm1, %%mm0 \n\t" // mm0 = differnece + "paddb %%mm7, %%mm0 \n\t" + "pcmpgtb %%mm6, %%mm0 \n\t" + + "movq (%%"REG_a",%3), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a", %3, 2), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a", %3, 2), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - - "lea (%%"REG_a", %3, 4), %%"REG_a" \n\t" - - "movq (%2, %3, 4), %%mm2 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + + "lea (%%"REG_a", %3, 4), %%"REG_a" \n\t" + + "movq (%2, %3, 4), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a"), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a"), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - - "movq (%%"REG_a", %3), %%mm2 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + + "movq (%%"REG_a", %3), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a", %3, 2), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a", %3, 2), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - "psubusb %%mm3, %%mm4 \n\t" - - " \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + "psubusb %%mm3, %%mm4 \n\t" + + " \n\t" #ifdef HAVE_MMX2 - "pxor %%mm7, %%mm7 \n\t" - "psadbw %%mm7, %%mm0 \n\t" + "pxor %%mm7, %%mm7 \n\t" + "psadbw %%mm7, %%mm0 \n\t" #else - "movq %%mm0, %%mm1 \n\t" - "psrlw $8, %%mm0 \n\t" - "paddb %%mm1, %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "psrlq $16, %%mm0 \n\t" - "paddb %%mm1, %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "psrlq $32, %%mm0 \n\t" - "paddb %%mm1, %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "psrlw $8, %%mm0 \n\t" + "paddb %%mm1, %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "psrlq $16, %%mm0 \n\t" + "paddb %%mm1, %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "psrlq $32, %%mm0 \n\t" + "paddb %%mm1, %%mm0 \n\t" #endif - "movq %4, %%mm7 \n\t" // QP,..., QP - "paddusb %%mm7, %%mm7 \n\t" // 2QP ... 2QP - "psubusb %%mm7, %%mm4 \n\t" // Diff <= 2QP -> 0 - "packssdw %%mm4, %%mm4 \n\t" - "movd %%mm0, %0 \n\t" - "movd %%mm4, %1 \n\t" - - : "=r" (numEq), "=r" (dcOk) - : "r" (src), "r" ((long)stride), "m" (c->pQPb) - : "%"REG_a - ); - - numEq= (-numEq) &0xFF; - if(numEq > c->ppMode.flatnessThreshold){ + "movq %4, %%mm7 \n\t" // QP,..., QP + "paddusb %%mm7, %%mm7 \n\t" // 2QP ... 2QP + "psubusb %%mm7, %%mm4 \n\t" // Diff <= 2QP -> 0 + "packssdw %%mm4, %%mm4 \n\t" + "movd %%mm0, %0 \n\t" + "movd %%mm4, %1 \n\t" + + : "=r" (numEq), "=r" (dcOk) + : "r" (src), "r" ((long)stride), "m" (c->pQPb) + : "%"REG_a + ); + + numEq= (-numEq) &0xFF; + if(numEq > c->ppMode.flatnessThreshold){ if(dcOk) return 0; else return 1; }else{ @@ -196,172 +196,172 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, PPContext *c) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*3; - asm volatile( //"movv %0 %1 %2\n\t" - "movq %2, %%mm0 \n\t" // QP,..., QP - "pxor %%mm4, %%mm4 \n\t" - - "movq (%0), %%mm6 \n\t" - "movq (%0, %1), %%mm5 \n\t" - "movq %%mm5, %%mm1 \n\t" - "movq %%mm6, %%mm2 \n\t" - "psubusb %%mm6, %%mm5 \n\t" - "psubusb %%mm1, %%mm2 \n\t" - "por %%mm5, %%mm2 \n\t" // ABS Diff of lines - "psubusb %%mm0, %%mm2 \n\t" // diff <= QP -> 0 - "pcmpeqb %%mm4, %%mm2 \n\t" // diff <= QP -> FF - - "pand %%mm2, %%mm6 \n\t" - "pandn %%mm1, %%mm2 \n\t" - "por %%mm2, %%mm6 \n\t"// First Line to Filter - - "movq (%0, %1, 8), %%mm5 \n\t" - "lea (%0, %1, 4), %%"REG_a" \n\t" - "lea (%0, %1, 8), %%"REG_c" \n\t" - "sub %1, %%"REG_c" \n\t" - "add %1, %0 \n\t" // %0 points to line 1 not 0 - "movq (%0, %1, 8), %%mm7 \n\t" - "movq %%mm5, %%mm1 \n\t" - "movq %%mm7, %%mm2 \n\t" - "psubusb %%mm7, %%mm5 \n\t" - "psubusb %%mm1, %%mm2 \n\t" - "por %%mm5, %%mm2 \n\t" // ABS Diff of lines - "psubusb %%mm0, %%mm2 \n\t" // diff <= QP -> 0 - "pcmpeqb %%mm4, %%mm2 \n\t" // diff <= QP -> FF - - "pand %%mm2, %%mm7 \n\t" - "pandn %%mm1, %%mm2 \n\t" - "por %%mm2, %%mm7 \n\t" // First Line to Filter - - - // 1 2 3 4 5 6 7 8 - // %0 %0+%1 %0+2%1 eax %0+4%1 eax+2%1 ecx eax+4%1 - // 6 4 2 2 1 1 - // 6 4 4 2 - // 6 8 2 - - "movq (%0, %1), %%mm0 \n\t" // 1 - "movq %%mm0, %%mm1 \n\t" // 1 - PAVGB(%%mm6, %%mm0) //1 1 /2 - PAVGB(%%mm6, %%mm0) //3 1 /4 - - "movq (%0, %1, 4), %%mm2 \n\t" // 1 - "movq %%mm2, %%mm5 \n\t" // 1 - PAVGB((%%REGa), %%mm2) // 11 /2 - PAVGB((%0, %1, 2), %%mm2) // 211 /4 - "movq %%mm2, %%mm3 \n\t" // 211 /4 - "movq (%0), %%mm4 \n\t" // 1 - PAVGB(%%mm4, %%mm3) // 4 211 /8 - PAVGB(%%mm0, %%mm3) //642211 /16 - "movq %%mm3, (%0) \n\t" // X - // mm1=2 mm2=3(211) mm4=1 mm5=5 mm6=0 mm7=9 - "movq %%mm1, %%mm0 \n\t" // 1 - PAVGB(%%mm6, %%mm0) //1 1 /2 - "movq %%mm4, %%mm3 \n\t" // 1 - PAVGB((%0,%1,2), %%mm3) // 1 1 /2 - PAVGB((%%REGa,%1,2), %%mm5) // 11 /2 - PAVGB((%%REGa), %%mm5) // 211 /4 - PAVGB(%%mm5, %%mm3) // 2 2211 /8 - PAVGB(%%mm0, %%mm3) //4242211 /16 - "movq %%mm3, (%0,%1) \n\t" // X - // mm1=2 mm2=3(211) mm4=1 mm5=4(211) mm6=0 mm7=9 - PAVGB(%%mm4, %%mm6) //11 /2 - "movq (%%"REG_c"), %%mm0 \n\t" // 1 - PAVGB((%%REGa, %1, 2), %%mm0) // 11/2 - "movq %%mm0, %%mm3 \n\t" // 11/2 - PAVGB(%%mm1, %%mm0) // 2 11/4 - PAVGB(%%mm6, %%mm0) //222 11/8 - PAVGB(%%mm2, %%mm0) //22242211/16 - "movq (%0, %1, 2), %%mm2 \n\t" // 1 - "movq %%mm0, (%0, %1, 2) \n\t" // X - // mm1=2 mm2=3 mm3=6(11) mm4=1 mm5=4(211) mm6=0(11) mm7=9 - "movq (%%"REG_a", %1, 4), %%mm0 \n\t" // 1 - PAVGB((%%REGc), %%mm0) // 11 /2 - PAVGB(%%mm0, %%mm6) //11 11 /4 - PAVGB(%%mm1, %%mm4) // 11 /2 - PAVGB(%%mm2, %%mm1) // 11 /2 - PAVGB(%%mm1, %%mm6) //1122 11 /8 - PAVGB(%%mm5, %%mm6) //112242211 /16 - "movq (%%"REG_a"), %%mm5 \n\t" // 1 - "movq %%mm6, (%%"REG_a") \n\t" // X - // mm0=7(11) mm1=2(11) mm2=3 mm3=6(11) mm4=1(11) mm5=4 mm7=9 - "movq (%%"REG_a", %1, 4), %%mm6 \n\t" // 1 - PAVGB(%%mm7, %%mm6) // 11 /2 - PAVGB(%%mm4, %%mm6) // 11 11 /4 - PAVGB(%%mm3, %%mm6) // 11 2211 /8 - PAVGB(%%mm5, %%mm2) // 11 /2 - "movq (%0, %1, 4), %%mm4 \n\t" // 1 - PAVGB(%%mm4, %%mm2) // 112 /4 - PAVGB(%%mm2, %%mm6) // 112242211 /16 - "movq %%mm6, (%0, %1, 4) \n\t" // X - // mm0=7(11) mm1=2(11) mm2=3(112) mm3=6(11) mm4=5 mm5=4 mm7=9 - PAVGB(%%mm7, %%mm1) // 11 2 /4 - PAVGB(%%mm4, %%mm5) // 11 /2 - PAVGB(%%mm5, %%mm0) // 11 11 /4 - "movq (%%"REG_a", %1, 2), %%mm6 \n\t" // 1 - PAVGB(%%mm6, %%mm1) // 11 4 2 /8 - PAVGB(%%mm0, %%mm1) // 11224222 /16 - "movq %%mm1, (%%"REG_a", %1, 2) \n\t" // X - // mm2=3(112) mm3=6(11) mm4=5 mm5=4(11) mm6=6 mm7=9 - PAVGB((%%REGc), %%mm2) // 112 4 /8 - "movq (%%"REG_a", %1, 4), %%mm0 \n\t" // 1 - PAVGB(%%mm0, %%mm6) // 1 1 /2 - PAVGB(%%mm7, %%mm6) // 1 12 /4 - PAVGB(%%mm2, %%mm6) // 1122424 /4 - "movq %%mm6, (%%"REG_c") \n\t" // X - // mm0=8 mm3=6(11) mm4=5 mm5=4(11) mm7=9 - PAVGB(%%mm7, %%mm5) // 11 2 /4 - PAVGB(%%mm7, %%mm5) // 11 6 /8 - - PAVGB(%%mm3, %%mm0) // 112 /4 - PAVGB(%%mm0, %%mm5) // 112246 /16 - "movq %%mm5, (%%"REG_a", %1, 4) \n\t" // X - "sub %1, %0 \n\t" - - : - : "r" (src), "r" ((long)stride), "m" (c->pQPb) - : "%"REG_a, "%"REG_c - ); + src+= stride*3; + asm volatile( //"movv %0 %1 %2\n\t" + "movq %2, %%mm0 \n\t" // QP,..., QP + "pxor %%mm4, %%mm4 \n\t" + + "movq (%0), %%mm6 \n\t" + "movq (%0, %1), %%mm5 \n\t" + "movq %%mm5, %%mm1 \n\t" + "movq %%mm6, %%mm2 \n\t" + "psubusb %%mm6, %%mm5 \n\t" + "psubusb %%mm1, %%mm2 \n\t" + "por %%mm5, %%mm2 \n\t" // ABS Diff of lines + "psubusb %%mm0, %%mm2 \n\t" // diff <= QP -> 0 + "pcmpeqb %%mm4, %%mm2 \n\t" // diff <= QP -> FF + + "pand %%mm2, %%mm6 \n\t" + "pandn %%mm1, %%mm2 \n\t" + "por %%mm2, %%mm6 \n\t"// First Line to Filter + + "movq (%0, %1, 8), %%mm5 \n\t" + "lea (%0, %1, 4), %%"REG_a" \n\t" + "lea (%0, %1, 8), %%"REG_c" \n\t" + "sub %1, %%"REG_c" \n\t" + "add %1, %0 \n\t" // %0 points to line 1 not 0 + "movq (%0, %1, 8), %%mm7 \n\t" + "movq %%mm5, %%mm1 \n\t" + "movq %%mm7, %%mm2 \n\t" + "psubusb %%mm7, %%mm5 \n\t" + "psubusb %%mm1, %%mm2 \n\t" + "por %%mm5, %%mm2 \n\t" // ABS Diff of lines + "psubusb %%mm0, %%mm2 \n\t" // diff <= QP -> 0 + "pcmpeqb %%mm4, %%mm2 \n\t" // diff <= QP -> FF + + "pand %%mm2, %%mm7 \n\t" + "pandn %%mm1, %%mm2 \n\t" + "por %%mm2, %%mm7 \n\t" // First Line to Filter + + + // 1 2 3 4 5 6 7 8 + // %0 %0+%1 %0+2%1 eax %0+4%1 eax+2%1 ecx eax+4%1 + // 6 4 2 2 1 1 + // 6 4 4 2 + // 6 8 2 + + "movq (%0, %1), %%mm0 \n\t" // 1 + "movq %%mm0, %%mm1 \n\t" // 1 + PAVGB(%%mm6, %%mm0) //1 1 /2 + PAVGB(%%mm6, %%mm0) //3 1 /4 + + "movq (%0, %1, 4), %%mm2 \n\t" // 1 + "movq %%mm2, %%mm5 \n\t" // 1 + PAVGB((%%REGa), %%mm2) // 11 /2 + PAVGB((%0, %1, 2), %%mm2) // 211 /4 + "movq %%mm2, %%mm3 \n\t" // 211 /4 + "movq (%0), %%mm4 \n\t" // 1 + PAVGB(%%mm4, %%mm3) // 4 211 /8 + PAVGB(%%mm0, %%mm3) //642211 /16 + "movq %%mm3, (%0) \n\t" // X + // mm1=2 mm2=3(211) mm4=1 mm5=5 mm6=0 mm7=9 + "movq %%mm1, %%mm0 \n\t" // 1 + PAVGB(%%mm6, %%mm0) //1 1 /2 + "movq %%mm4, %%mm3 \n\t" // 1 + PAVGB((%0,%1,2), %%mm3) // 1 1 /2 + PAVGB((%%REGa,%1,2), %%mm5) // 11 /2 + PAVGB((%%REGa), %%mm5) // 211 /4 + PAVGB(%%mm5, %%mm3) // 2 2211 /8 + PAVGB(%%mm0, %%mm3) //4242211 /16 + "movq %%mm3, (%0,%1) \n\t" // X + // mm1=2 mm2=3(211) mm4=1 mm5=4(211) mm6=0 mm7=9 + PAVGB(%%mm4, %%mm6) //11 /2 + "movq (%%"REG_c"), %%mm0 \n\t" // 1 + PAVGB((%%REGa, %1, 2), %%mm0) // 11/2 + "movq %%mm0, %%mm3 \n\t" // 11/2 + PAVGB(%%mm1, %%mm0) // 2 11/4 + PAVGB(%%mm6, %%mm0) //222 11/8 + PAVGB(%%mm2, %%mm0) //22242211/16 + "movq (%0, %1, 2), %%mm2 \n\t" // 1 + "movq %%mm0, (%0, %1, 2) \n\t" // X + // mm1=2 mm2=3 mm3=6(11) mm4=1 mm5=4(211) mm6=0(11) mm7=9 + "movq (%%"REG_a", %1, 4), %%mm0 \n\t" // 1 + PAVGB((%%REGc), %%mm0) // 11 /2 + PAVGB(%%mm0, %%mm6) //11 11 /4 + PAVGB(%%mm1, %%mm4) // 11 /2 + PAVGB(%%mm2, %%mm1) // 11 /2 + PAVGB(%%mm1, %%mm6) //1122 11 /8 + PAVGB(%%mm5, %%mm6) //112242211 /16 + "movq (%%"REG_a"), %%mm5 \n\t" // 1 + "movq %%mm6, (%%"REG_a") \n\t" // X + // mm0=7(11) mm1=2(11) mm2=3 mm3=6(11) mm4=1(11) mm5=4 mm7=9 + "movq (%%"REG_a", %1, 4), %%mm6 \n\t" // 1 + PAVGB(%%mm7, %%mm6) // 11 /2 + PAVGB(%%mm4, %%mm6) // 11 11 /4 + PAVGB(%%mm3, %%mm6) // 11 2211 /8 + PAVGB(%%mm5, %%mm2) // 11 /2 + "movq (%0, %1, 4), %%mm4 \n\t" // 1 + PAVGB(%%mm4, %%mm2) // 112 /4 + PAVGB(%%mm2, %%mm6) // 112242211 /16 + "movq %%mm6, (%0, %1, 4) \n\t" // X + // mm0=7(11) mm1=2(11) mm2=3(112) mm3=6(11) mm4=5 mm5=4 mm7=9 + PAVGB(%%mm7, %%mm1) // 11 2 /4 + PAVGB(%%mm4, %%mm5) // 11 /2 + PAVGB(%%mm5, %%mm0) // 11 11 /4 + "movq (%%"REG_a", %1, 2), %%mm6 \n\t" // 1 + PAVGB(%%mm6, %%mm1) // 11 4 2 /8 + PAVGB(%%mm0, %%mm1) // 11224222 /16 + "movq %%mm1, (%%"REG_a", %1, 2) \n\t" // X + // mm2=3(112) mm3=6(11) mm4=5 mm5=4(11) mm6=6 mm7=9 + PAVGB((%%REGc), %%mm2) // 112 4 /8 + "movq (%%"REG_a", %1, 4), %%mm0 \n\t" // 1 + PAVGB(%%mm0, %%mm6) // 1 1 /2 + PAVGB(%%mm7, %%mm6) // 1 12 /4 + PAVGB(%%mm2, %%mm6) // 1122424 /4 + "movq %%mm6, (%%"REG_c") \n\t" // X + // mm0=8 mm3=6(11) mm4=5 mm5=4(11) mm7=9 + PAVGB(%%mm7, %%mm5) // 11 2 /4 + PAVGB(%%mm7, %%mm5) // 11 6 /8 + + PAVGB(%%mm3, %%mm0) // 112 /4 + PAVGB(%%mm0, %%mm5) // 112246 /16 + "movq %%mm5, (%%"REG_a", %1, 4) \n\t" // X + "sub %1, %0 \n\t" + + : + : "r" (src), "r" ((long)stride), "m" (c->pQPb) + : "%"REG_a, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - const int l1= stride; - const int l2= stride + l1; - const int l3= stride + l2; - const int l4= stride + l3; - const int l5= stride + l4; - const int l6= stride + l5; - const int l7= stride + l6; - const int l8= stride + l7; - const int l9= stride + l8; - int x; - src+= stride*3; - for(x=0; x<BLOCK_SIZE; x++) - { - const int first= ABS(src[0] - src[l1]) < c->QP ? src[0] : src[l1]; - const int last= ABS(src[l8] - src[l9]) < c->QP ? src[l9] : src[l8]; - - int sums[10]; - sums[0] = 4*first + src[l1] + src[l2] + src[l3] + 4; - sums[1] = sums[0] - first + src[l4]; - sums[2] = sums[1] - first + src[l5]; - sums[3] = sums[2] - first + src[l6]; - sums[4] = sums[3] - first + src[l7]; - sums[5] = sums[4] - src[l1] + src[l8]; - sums[6] = sums[5] - src[l2] + last; - sums[7] = sums[6] - src[l3] + last; - sums[8] = sums[7] - src[l4] + last; - sums[9] = sums[8] - src[l5] + last; - - src[l1]= (sums[0] + sums[2] + 2*src[l1])>>4; - src[l2]= (sums[1] + sums[3] + 2*src[l2])>>4; - src[l3]= (sums[2] + sums[4] + 2*src[l3])>>4; - src[l4]= (sums[3] + sums[5] + 2*src[l4])>>4; - src[l5]= (sums[4] + sums[6] + 2*src[l5])>>4; - src[l6]= (sums[5] + sums[7] + 2*src[l6])>>4; - src[l7]= (sums[6] + sums[8] + 2*src[l7])>>4; - src[l8]= (sums[7] + sums[9] + 2*src[l8])>>4; - - src++; - } + const int l1= stride; + const int l2= stride + l1; + const int l3= stride + l2; + const int l4= stride + l3; + const int l5= stride + l4; + const int l6= stride + l5; + const int l7= stride + l6; + const int l8= stride + l7; + const int l9= stride + l8; + int x; + src+= stride*3; + for(x=0; x<BLOCK_SIZE; x++) + { + const int first= ABS(src[0] - src[l1]) < c->QP ? src[0] : src[l1]; + const int last= ABS(src[l8] - src[l9]) < c->QP ? src[l9] : src[l8]; + + int sums[10]; + sums[0] = 4*first + src[l1] + src[l2] + src[l3] + 4; + sums[1] = sums[0] - first + src[l4]; + sums[2] = sums[1] - first + src[l5]; + sums[3] = sums[2] - first + src[l6]; + sums[4] = sums[3] - first + src[l7]; + sums[5] = sums[4] - src[l1] + src[l8]; + sums[6] = sums[5] - src[l2] + last; + sums[7] = sums[6] - src[l3] + last; + sums[8] = sums[7] - src[l4] + last; + sums[9] = sums[8] - src[l5] + last; + + src[l1]= (sums[0] + sums[2] + 2*src[l1])>>4; + src[l2]= (sums[1] + sums[3] + 2*src[l2])>>4; + src[l3]= (sums[2] + sums[4] + 2*src[l3])>>4; + src[l4]= (sums[3] + sums[5] + 2*src[l4])>>4; + src[l5]= (sums[4] + sums[6] + 2*src[l5])>>4; + src[l6]= (sums[5] + sums[7] + 2*src[l6])>>4; + src[l7]= (sums[6] + sums[8] + 2*src[l7])>>4; + src[l8]= (sums[7] + sums[9] + 2*src[l8])>>4; + + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } #endif //HAVE_ALTIVEC @@ -372,101 +372,101 @@ * values are correctly clipped (MMX2) * values are wraparound (C) * conclusion: its fast, but introduces ugly horizontal patterns if there is a continious gradient - 0 8 16 24 - x = 8 - x/2 = 4 - x/8 = 1 - 1 12 12 23 + 0 8 16 24 + x = 8 + x/2 = 4 + x/8 = 1 + 1 12 12 23 */ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*3; + src+= stride*3; // FIXME rounding - asm volatile( - "pxor %%mm7, %%mm7 \n\t" // 0 - "movq "MANGLE(b80)", %%mm6 \n\t" // MIN_SIGNED_BYTE - "leal (%0, %1), %%"REG_a" \n\t" - "leal (%%"REG_a", %1, 4), %%"REG_c" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 - "movq "MANGLE(pQPb)", %%mm0 \n\t" // QP,..., QP - "movq %%mm0, %%mm1 \n\t" // QP,..., QP - "paddusb "MANGLE(b02)", %%mm0 \n\t" - "psrlw $2, %%mm0 \n\t" - "pand "MANGLE(b3F)", %%mm0 \n\t" // QP/4,..., QP/4 - "paddusb %%mm1, %%mm0 \n\t" // QP*1.25 ... - "movq (%0, %1, 4), %%mm2 \n\t" // line 4 - "movq (%%"REG_c"), %%mm3 \n\t" // line 5 - "movq %%mm2, %%mm4 \n\t" // line 4 - "pcmpeqb %%mm5, %%mm5 \n\t" // -1 - "pxor %%mm2, %%mm5 \n\t" // -line 4 - 1 - PAVGB(%%mm3, %%mm5) - "paddb %%mm6, %%mm5 \n\t" // (l5-l4)/2 - "psubusb %%mm3, %%mm4 \n\t" - "psubusb %%mm2, %%mm3 \n\t" - "por %%mm3, %%mm4 \n\t" // |l4 - l5| - "psubusb %%mm0, %%mm4 \n\t" - "pcmpeqb %%mm7, %%mm4 \n\t" - "pand %%mm4, %%mm5 \n\t" // d/2 - -// "paddb %%mm6, %%mm2 \n\t" // line 4 + 0x80 - "paddb %%mm5, %%mm2 \n\t" -// "psubb %%mm6, %%mm2 \n\t" - "movq %%mm2, (%0,%1, 4) \n\t" - - "movq (%%"REG_c"), %%mm2 \n\t" -// "paddb %%mm6, %%mm2 \n\t" // line 5 + 0x80 - "psubb %%mm5, %%mm2 \n\t" -// "psubb %%mm6, %%mm2 \n\t" - "movq %%mm2, (%%"REG_c") \n\t" - - "paddb %%mm6, %%mm5 \n\t" - "psrlw $2, %%mm5 \n\t" - "pand "MANGLE(b3F)", %%mm5 \n\t" - "psubb "MANGLE(b20)", %%mm5 \n\t" // (l5-l4)/8 - - "movq (%%"REG_a", %1, 2), %%mm2 \n\t" - "paddb %%mm6, %%mm2 \n\t" // line 3 + 0x80 - "paddsb %%mm5, %%mm2 \n\t" - "psubb %%mm6, %%mm2 \n\t" - "movq %%mm2, (%%"REG_a", %1, 2) \n\t" - - "movq (%%"REG_c", %1), %%mm2 \n\t" - "paddb %%mm6, %%mm2 \n\t" // line 6 + 0x80 - "psubsb %%mm5, %%mm2 \n\t" - "psubb %%mm6, %%mm2 \n\t" - "movq %%mm2, (%%"REG_c", %1) \n\t" - - : - : "r" (src), "r" ((long)stride) - : "%"REG_a, "%"REG_c - ); + asm volatile( + "pxor %%mm7, %%mm7 \n\t" // 0 + "movq "MANGLE(b80)", %%mm6 \n\t" // MIN_SIGNED_BYTE + "leal (%0, %1), %%"REG_a" \n\t" + "leal (%%"REG_a", %1, 4), %%"REG_c" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 + "movq "MANGLE(pQPb)", %%mm0 \n\t" // QP,..., QP + "movq %%mm0, %%mm1 \n\t" // QP,..., QP + "paddusb "MANGLE(b02)", %%mm0 \n\t" + "psrlw $2, %%mm0 \n\t" + "pand "MANGLE(b3F)", %%mm0 \n\t" // QP/4,..., QP/4 + "paddusb %%mm1, %%mm0 \n\t" // QP*1.25 ... + "movq (%0, %1, 4), %%mm2 \n\t" // line 4 + "movq (%%"REG_c"), %%mm3 \n\t" // line 5 + "movq %%mm2, %%mm4 \n\t" // line 4 + "pcmpeqb %%mm5, %%mm5 \n\t" // -1 + "pxor %%mm2, %%mm5 \n\t" // -line 4 - 1 + PAVGB(%%mm3, %%mm5) + "paddb %%mm6, %%mm5 \n\t" // (l5-l4)/2 + "psubusb %%mm3, %%mm4 \n\t" + "psubusb %%mm2, %%mm3 \n\t" + "por %%mm3, %%mm4 \n\t" // |l4 - l5| + "psubusb %%mm0, %%mm4 \n\t" + "pcmpeqb %%mm7, %%mm4 \n\t" + "pand %%mm4, %%mm5 \n\t" // d/2 + +// "paddb %%mm6, %%mm2 \n\t" // line 4 + 0x80 + "paddb %%mm5, %%mm2 \n\t" +// "psubb %%mm6, %%mm2 \n\t" + "movq %%mm2, (%0,%1, 4) \n\t" + + "movq (%%"REG_c"), %%mm2 \n\t" +// "paddb %%mm6, %%mm2 \n\t" // line 5 + 0x80 + "psubb %%mm5, %%mm2 \n\t" +// "psubb %%mm6, %%mm2 \n\t" + "movq %%mm2, (%%"REG_c") \n\t" + + "paddb %%mm6, %%mm5 \n\t" + "psrlw $2, %%mm5 \n\t" + "pand "MANGLE(b3F)", %%mm5 \n\t" + "psubb "MANGLE(b20)", %%mm5 \n\t" // (l5-l4)/8 + + "movq (%%"REG_a", %1, 2), %%mm2 \n\t" + "paddb %%mm6, %%mm2 \n\t" // line 3 + 0x80 + "paddsb %%mm5, %%mm2 \n\t" + "psubb %%mm6, %%mm2 \n\t" + "movq %%mm2, (%%"REG_a", %1, 2) \n\t" + + "movq (%%"REG_c", %1), %%mm2 \n\t" + "paddb %%mm6, %%mm2 \n\t" // line 6 + 0x80 + "psubsb %%mm5, %%mm2 \n\t" + "psubb %%mm6, %%mm2 \n\t" + "movq %%mm2, (%%"REG_c", %1) \n\t" + + : + : "r" (src), "r" ((long)stride) + : "%"REG_a, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - const int l1= stride; - const int l2= stride + l1; - const int l3= stride + l2; - const int l4= stride + l3; - const int l5= stride + l4; - const int l6= stride + l5; -// const int l7= stride + l6; -// const int l8= stride + l7; -// const int l9= stride + l8; - int x; - const int QP15= QP + (QP>>2); - src+= stride*3; - for(x=0; x<BLOCK_SIZE; x++) - { - const int v = (src[x+l5] - src[x+l4]); - if(ABS(v) < QP15) - { - src[x+l3] +=v>>3; - src[x+l4] +=v>>1; - src[x+l5] -=v>>1; - src[x+l6] -=v>>3; - - } - } + const int l1= stride; + const int l2= stride + l1; + const int l3= stride + l2; + const int l4= stride + l3; + const int l5= stride + l4; + const int l6= stride + l5; +// const int l7= stride + l6; +// const int l8= stride + l7; +// const int l9= stride + l8; + int x; + const int QP15= QP + (QP>>2); + src+= stride*3; + for(x=0; x<BLOCK_SIZE; x++) + { + const int v = (src[x+l5] - src[x+l4]); + if(ABS(v) < QP15) + { + src[x+l3] +=v>>3; + src[x+l4] +=v>>1; + src[x+l5] -=v>>1; + src[x+l6] -=v>>3; + + } + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -482,128 +482,128 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, PPContext *co) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*3; - - asm volatile( - "pxor %%mm7, %%mm7 \n\t" // 0 - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" // line 3 - "movq (%0, %1, 4), %%mm1 \n\t" // line 4 - "movq %%mm1, %%mm2 \n\t" // line 4 - "psubusb %%mm0, %%mm1 \n\t" - "psubusb %%mm2, %%mm0 \n\t" - "por %%mm1, %%mm0 \n\t" // |l2 - l3| - "movq (%%"REG_c"), %%mm3 \n\t" // line 5 - "movq (%%"REG_c", %1), %%mm4 \n\t" // line 6 - "movq %%mm3, %%mm5 \n\t" // line 5 - "psubusb %%mm4, %%mm3 \n\t" - "psubusb %%mm5, %%mm4 \n\t" - "por %%mm4, %%mm3 \n\t" // |l5 - l6| - PAVGB(%%mm3, %%mm0) // (|l2 - l3| + |l5 - l6|)/2 - "movq %%mm2, %%mm1 \n\t" // line 4 - "psubusb %%mm5, %%mm2 \n\t" - "movq %%mm2, %%mm4 \n\t" - "pcmpeqb %%mm7, %%mm2 \n\t" // (l4 - l5) <= 0 ? -1 : 0 - "psubusb %%mm1, %%mm5 \n\t" - "por %%mm5, %%mm4 \n\t" // |l4 - l5| - "psubusb %%mm0, %%mm4 \n\t" //d = MAX(0, |l4-l5| - (|l2-l3| + |l5-l6|)/2) - "movq %%mm4, %%mm3 \n\t" // d - "movq %2, %%mm0 \n\t" - "paddusb %%mm0, %%mm0 \n\t" - "psubusb %%mm0, %%mm4 \n\t" - "pcmpeqb %%mm7, %%mm4 \n\t" // d <= QP ? -1 : 0 - "psubusb "MANGLE(b01)", %%mm3 \n\t" - "pand %%mm4, %%mm3 \n\t" // d <= QP ? d : 0 - - PAVGB(%%mm7, %%mm3) // d/2 - "movq %%mm3, %%mm1 \n\t" // d/2 - PAVGB(%%mm7, %%mm3) // d/4 - PAVGB(%%mm1, %%mm3) // 3*d/8 - - "movq (%0, %1, 4), %%mm0 \n\t" // line 4 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l4-1 : l4 - "psubusb %%mm3, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%0, %1, 4) \n\t" // line 4 - - "movq (%%"REG_c"), %%mm0 \n\t" // line 5 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5 - "paddusb %%mm3, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%%"REG_c") \n\t" // line 5 - - PAVGB(%%mm7, %%mm1) // d/4 - - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" // line 3 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l4-1 : l4 - "psubusb %%mm1, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%%"REG_a", %1, 2) \n\t" // line 3 - - "movq (%%"REG_c", %1), %%mm0 \n\t" // line 6 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5 - "paddusb %%mm1, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%%"REG_c", %1) \n\t" // line 6 - - PAVGB(%%mm7, %%mm1) // d/8 - - "movq (%%"REG_a", %1), %%mm0 \n\t" // line 2 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l2-1 : l2 - "psubusb %%mm1, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%%"REG_a", %1) \n\t" // line 2 - - "movq (%%"REG_c", %1, 2), %%mm0 \n\t" // line 7 - "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l7-1 : l7 - "paddusb %%mm1, %%mm0 \n\t" - "pxor %%mm2, %%mm0 \n\t" - "movq %%mm0, (%%"REG_c", %1, 2) \n\t" // line 7 - - : - : "r" (src), "r" ((long)stride), "m" (co->pQPb) - : "%"REG_a, "%"REG_c - ); + src+= stride*3; + + asm volatile( + "pxor %%mm7, %%mm7 \n\t" // 0 + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" // line 3 + "movq (%0, %1, 4), %%mm1 \n\t" // line 4 + "movq %%mm1, %%mm2 \n\t" // line 4 + "psubusb %%mm0, %%mm1 \n\t" + "psubusb %%mm2, %%mm0 \n\t" + "por %%mm1, %%mm0 \n\t" // |l2 - l3| + "movq (%%"REG_c"), %%mm3 \n\t" // line 5 + "movq (%%"REG_c", %1), %%mm4 \n\t" // line 6 + "movq %%mm3, %%mm5 \n\t" // line 5 + "psubusb %%mm4, %%mm3 \n\t" + "psubusb %%mm5, %%mm4 \n\t" + "por %%mm4, %%mm3 \n\t" // |l5 - l6| + PAVGB(%%mm3, %%mm0) // (|l2 - l3| + |l5 - l6|)/2 + "movq %%mm2, %%mm1 \n\t" // line 4 + "psubusb %%mm5, %%mm2 \n\t" + "movq %%mm2, %%mm4 \n\t" + "pcmpeqb %%mm7, %%mm2 \n\t" // (l4 - l5) <= 0 ? -1 : 0 + "psubusb %%mm1, %%mm5 \n\t" + "por %%mm5, %%mm4 \n\t" // |l4 - l5| + "psubusb %%mm0, %%mm4 \n\t" //d = MAX(0, |l4-l5| - (|l2-l3| + |l5-l6|)/2) + "movq %%mm4, %%mm3 \n\t" // d + "movq %2, %%mm0 \n\t" + "paddusb %%mm0, %%mm0 \n\t" + "psubusb %%mm0, %%mm4 \n\t" + "pcmpeqb %%mm7, %%mm4 \n\t" // d <= QP ? -1 : 0 + "psubusb "MANGLE(b01)", %%mm3 \n\t" + "pand %%mm4, %%mm3 \n\t" // d <= QP ? d : 0 + + PAVGB(%%mm7, %%mm3) // d/2 + "movq %%mm3, %%mm1 \n\t" // d/2 + PAVGB(%%mm7, %%mm3) // d/4 + PAVGB(%%mm1, %%mm3) // 3*d/8 + + "movq (%0, %1, 4), %%mm0 \n\t" // line 4 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l4-1 : l4 + "psubusb %%mm3, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%0, %1, 4) \n\t" // line 4 + + "movq (%%"REG_c"), %%mm0 \n\t" // line 5 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5 + "paddusb %%mm3, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%%"REG_c") \n\t" // line 5 + + PAVGB(%%mm7, %%mm1) // d/4 + + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" // line 3 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l4-1 : l4 + "psubusb %%mm1, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%%"REG_a", %1, 2) \n\t" // line 3 + + "movq (%%"REG_c", %1), %%mm0 \n\t" // line 6 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5 + "paddusb %%mm1, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%%"REG_c", %1) \n\t" // line 6 + + PAVGB(%%mm7, %%mm1) // d/8 + + "movq (%%"REG_a", %1), %%mm0 \n\t" // line 2 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l2-1 : l2 + "psubusb %%mm1, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%%"REG_a", %1) \n\t" // line 2 + + "movq (%%"REG_c", %1, 2), %%mm0 \n\t" // line 7 + "pxor %%mm2, %%mm0 \n\t" //(l4 - l5) <= 0 ? -l7-1 : l7 + "paddusb %%mm1, %%mm0 \n\t" + "pxor %%mm2, %%mm0 \n\t" + "movq %%mm0, (%%"REG_c", %1, 2) \n\t" // line 7 + + : + : "r" (src), "r" ((long)stride), "m" (co->pQPb) + : "%"REG_a, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - const int l1= stride; - const int l2= stride + l1; - const int l3= stride + l2; - const int l4= stride + l3; - const int l5= stride + l4; - const int l6= stride + l5; - const int l7= stride + l6; -// const int l8= stride + l7; -// const int l9= stride + l8; - int x; - - src+= stride*3; - for(x=0; x<BLOCK_SIZE; x++) - { - int a= src[l3] - src[l4]; - int b= src[l4] - src[l5]; - int c= src[l5] - src[l6]; - - int d= ABS(b) - ((ABS(a) + ABS(c))>>1); - d= MAX(d, 0); - - if(d < co->QP*2) - { - int v = d * SIGN(-b); - - src[l2] +=v>>3; - src[l3] +=v>>2; - src[l4] +=(3*v)>>3; - src[l5] -=(3*v)>>3; - src[l6] -=v>>2; - src[l7] -=v>>3; - - } - src++; - } + const int l1= stride; + const int l2= stride + l1; + const int l3= stride + l2; + const int l4= stride + l3; + const int l5= stride + l4; + const int l6= stride + l5; + const int l7= stride + l6; +// const int l8= stride + l7; +// const int l9= stride + l8; + int x; + + src+= stride*3; + for(x=0; x<BLOCK_SIZE; x++) + { + int a= src[l3] - src[l4]; + int b= src[l4] - src[l5]; + int c= src[l5] - src[l6]; + + int d= ABS(b) - ((ABS(a) + ABS(c))>>1); + d= MAX(d, 0); + + if(d < co->QP*2) + { + int v = d * SIGN(-b); + + src[l2] +=v>>3; + src[l3] +=v>>2; + src[l4] +=(3*v)>>3; + src[l5] -=(3*v)>>3; + src[l6] -=v>>2; + src[l7] -=v>>3; + + } + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -612,569 +612,569 @@ { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) /* - uint8_t tmp[16]; - const int l1= stride; - const int l2= stride + l1; - const int l3= stride + l2; - const int l4= (int)tmp - (int)src - stride*3; - const int l5= (int)tmp - (int)src - stride*3 + 8; - const int l6= stride*3 + l3; - const int l7= stride + l6; - const int l8= stride + l7; - - memcpy(tmp, src+stride*7, 8); - memcpy(tmp+8, src+stride*8, 8); + uint8_t tmp[16]; + const int l1= stride; + const int l2= stride + l1; + const int l3= stride + l2; + const int l4= (int)tmp - (int)src - stride*3; + const int l5= (int)tmp - (int)src - stride*3 + 8; + const int l6= stride*3 + l3; + const int l7= stride + l6; + const int l8= stride + l7; + + memcpy(tmp, src+stride*7, 8); + memcpy(tmp+8, src+stride*8, 8); */ - src+= stride*4; - asm volatile( + src+= stride*4; + asm volatile( #if 0 //sligtly more accurate and slightly slower - "pxor %%mm7, %%mm7 \n\t" // 0 - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" -// 0 1 2 3 4 5 6 7 -// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 ecx+%1 ecx+2%1 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 - - - "movq (%0, %1, 2), %%mm0 \n\t" // l2 - "movq (%0), %%mm1 \n\t" // l0 - "movq %%mm0, %%mm2 \n\t" // l2 - PAVGB(%%mm7, %%mm0) // ~l2/2 - PAVGB(%%mm1, %%mm0) // ~(l2 + 2l0)/4 - PAVGB(%%mm2, %%mm0) // ~(5l2 + 2l0)/8 - - "movq (%%"REG_a"), %%mm1 \n\t" // l1 - "movq (%%"REG_a", %1, 2), %%mm3 \n\t" // l3 - "movq %%mm1, %%mm4 \n\t" // l1 - PAVGB(%%mm7, %%mm1) // ~l1/2 - PAVGB(%%mm3, %%mm1) // ~(l1 + 2l3)/4 - PAVGB(%%mm4, %%mm1) // ~(5l1 + 2l3)/8 - - "movq %%mm0, %%mm4 \n\t" // ~(5l2 + 2l0)/8 - "psubusb %%mm1, %%mm0 \n\t" - "psubusb %%mm4, %%mm1 \n\t" - "por %%mm0, %%mm1 \n\t" // ~|2l0 - 5l1 + 5l2 - 2l3|/8 + "pxor %%mm7, %%mm7 \n\t" // 0 + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" +// 0 1 2 3 4 5 6 7 +// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 ecx+%1 ecx+2%1 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 + + + "movq (%0, %1, 2), %%mm0 \n\t" // l2 + "movq (%0), %%mm1 \n\t" // l0 + "movq %%mm0, %%mm2 \n\t" // l2 + PAVGB(%%mm7, %%mm0) // ~l2/2 + PAVGB(%%mm1, %%mm0) // ~(l2 + 2l0)/4 + PAVGB(%%mm2, %%mm0) // ~(5l2 + 2l0)/8 + + "movq (%%"REG_a"), %%mm1 \n\t" // l1 + "movq (%%"REG_a", %1, 2), %%mm3 \n\t" // l3 + "movq %%mm1, %%mm4 \n\t" // l1 + PAVGB(%%mm7, %%mm1) // ~l1/2 + PAVGB(%%mm3, %%mm1) // ~(l1 + 2l3)/4 + PAVGB(%%mm4, %%mm1) // ~(5l1 + 2l3)/8 + + "movq %%mm0, %%mm4 \n\t" // ~(5l2 + 2l0)/8 + "psubusb %%mm1, %%mm0 \n\t" + "psubusb %%mm4, %%mm1 \n\t" + "por %%mm0, %%mm1 \n\t" // ~|2l0 - 5l1 + 5l2 - 2l3|/8 // mm1= |lenergy|, mm2= l2, mm3= l3, mm7=0 - "movq (%0, %1, 4), %%mm0 \n\t" // l4 - "movq %%mm0, %%mm4 \n\t" // l4 - PAVGB(%%mm7, %%mm0) // ~l4/2 - PAVGB(%%mm2, %%mm0) // ~(l4 + 2l2)/4 - PAVGB(%%mm4, %%mm0) // ~(5l4 + 2l2)/8 - - "movq (%%"REG_c"), %%mm2 \n\t" // l5 - "movq %%mm3, %%mm5 \n\t" // l3 - PAVGB(%%mm7, %%mm3) // ~l3/2 - PAVGB(%%mm2, %%mm3) // ~(l3 + 2l5)/4 - PAVGB(%%mm5, %%mm3) // ~(5l3 + 2l5)/8 - - "movq %%mm0, %%mm6 \n\t" // ~(5l4 + 2l2)/8 - "psubusb %%mm3, %%mm0 \n\t" - "psubusb %%mm6, %%mm3 \n\t" - "por %%mm0, %%mm3 \n\t" // ~|2l2 - 5l3 + 5l4 - 2l5|/8 - "pcmpeqb %%mm7, %%mm0 \n\t" // SIGN(2l2 - 5l3 + 5l4 - 2l5) + "movq (%0, %1, 4), %%mm0 \n\t" // l4 + "movq %%mm0, %%mm4 \n\t" // l4 + PAVGB(%%mm7, %%mm0) // ~l4/2 + PAVGB(%%mm2, %%mm0) // ~(l4 + 2l2)/4 + PAVGB(%%mm4, %%mm0) // ~(5l4 + 2l2)/8 + + "movq (%%"REG_c"), %%mm2 \n\t" // l5 + "movq %%mm3, %%mm5 \n\t" // l3 + PAVGB(%%mm7, %%mm3) // ~l3/2 + PAVGB(%%mm2, %%mm3) // ~(l3 + 2l5)/4 + PAVGB(%%mm5, %%mm3) // ~(5l3 + 2l5)/8 + + "movq %%mm0, %%mm6 \n\t" // ~(5l4 + 2l2)/8 + "psubusb %%mm3, %%mm0 \n\t" + "psubusb %%mm6, %%mm3 \n\t" + "por %%mm0, %%mm3 \n\t" // ~|2l2 - 5l3 + 5l4 - 2l5|/8 + "pcmpeqb %%mm7, %%mm0 \n\t" // SIGN(2l2 - 5l3 + 5l4 - 2l5) // mm0= SIGN(menergy), mm1= |lenergy|, mm2= l5, mm3= |menergy|, mm4=l4, mm5= l3, mm7=0 - "movq (%%"REG_c", %1), %%mm6 \n\t" // l6 - "movq %%mm6, %%mm5 \n\t" // l6 - PAVGB(%%mm7, %%mm6) // ~l6/2 - PAVGB(%%mm4, %%mm6) // ~(l6 + 2l4)/4 - PAVGB(%%mm5, %%mm6) // ~(5l6 + 2l4)/8 - - "movq (%%"REG_c", %1, 2), %%mm5 \n\t" // l7 - "movq %%mm2, %%mm4 \n\t" // l5 - PAVGB(%%mm7, %%mm2) // ~l5/2 - PAVGB(%%mm5, %%mm2) // ~(l5 + 2l7)/4 - PAVGB(%%mm4, %%mm2) // ~(5l5 + 2l7)/8 - - "movq %%mm6, %%mm4 \n\t" // ~(5l6 + 2l4)/8 - "psubusb %%mm2, %%mm6 \n\t" - "psubusb %%mm4, %%mm2 \n\t" - "por %%mm6, %%mm2 \n\t" // ~|2l4 - 5l5 + 5l6 - 2l7|/8 + "movq (%%"REG_c", %1), %%mm6 \n\t" // l6 + "movq %%mm6, %%mm5 \n\t" // l6 + PAVGB(%%mm7, %%mm6) // ~l6/2 + PAVGB(%%mm4, %%mm6) // ~(l6 + 2l4)/4 + PAVGB(%%mm5, %%mm6) // ~(5l6 + 2l4)/8 + + "movq (%%"REG_c", %1, 2), %%mm5 \n\t" // l7 + "movq %%mm2, %%mm4 \n\t" // l5 + PAVGB(%%mm7, %%mm2) // ~l5/2 + PAVGB(%%mm5, %%mm2) // ~(l5 + 2l7)/4 + PAVGB(%%mm4, %%mm2) // ~(5l5 + 2l7)/8 + + "movq %%mm6, %%mm4 \n\t" // ~(5l6 + 2l4)/8 + "psubusb %%mm2, %%mm6 \n\t" + "psubusb %%mm4, %%mm2 \n\t" + "por %%mm6, %%mm2 \n\t" // ~|2l4 - 5l5 + 5l6 - 2l7|/8 // mm0= SIGN(menergy), mm1= |lenergy|/8, mm2= |renergy|/8, mm3= |menergy|/8, mm7=0 - PMINUB(%%mm2, %%mm1, %%mm4) // MIN(|lenergy|,|renergy|)/8 - "movq %2, %%mm4 \n\t" // QP //FIXME QP+1 ? - "paddusb "MANGLE(b01)", %%mm4 \n\t" - "pcmpgtb %%mm3, %%mm4 \n\t" // |menergy|/8 < QP - "psubusb %%mm1, %%mm3 \n\t" // d=|menergy|/8-MIN(|lenergy|,|renergy|)/8 - "pand %%mm4, %%mm3 \n\t" - - "movq %%mm3, %%mm1 \n\t" -// "psubusb "MANGLE(b01)", %%mm3 \n\t" - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm7, %%mm3) - "paddusb %%mm1, %%mm3 \n\t" -// "paddusb "MANGLE(b01)", %%mm3 \n\t" - - "movq (%%"REG_a", %1, 2), %%mm6 \n\t" //l3 - "movq (%0, %1, 4), %%mm5 \n\t" //l4 - "movq (%0, %1, 4), %%mm4 \n\t" //l4 - "psubusb %%mm6, %%mm5 \n\t" - "psubusb %%mm4, %%mm6 \n\t" - "por %%mm6, %%mm5 \n\t" // |l3-l4| - "pcmpeqb %%mm7, %%mm6 \n\t" // SIGN(l3-l4) - "pxor %%mm6, %%mm0 \n\t" - "pand %%mm0, %%mm3 \n\t" - PMINUB(%%mm5, %%mm3, %%mm0) - - "psubusb "MANGLE(b01)", %%mm3 \n\t" - PAVGB(%%mm7, %%mm3) - - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" - "movq (%0, %1, 4), %%mm2 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "psubb %%mm3, %%mm0 \n\t" - "paddb %%mm3, %%mm2 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "movq %%mm0, (%%"REG_a", %1, 2) \n\t" - "movq %%mm2, (%0, %1, 4) \n\t" + PMINUB(%%mm2, %%mm1, %%mm4) // MIN(|lenergy|,|renergy|)/8 + "movq %2, %%mm4 \n\t" // QP //FIXME QP+1 ? + "paddusb "MANGLE(b01)", %%mm4 \n\t" + "pcmpgtb %%mm3, %%mm4 \n\t" // |menergy|/8 < QP + "psubusb %%mm1, %%mm3 \n\t" // d=|menergy|/8-MIN(|lenergy|,|renergy|)/8 + "pand %%mm4, %%mm3 \n\t" + + "movq %%mm3, %%mm1 \n\t" +// "psubusb "MANGLE(b01)", %%mm3 \n\t" + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm7, %%mm3) + "paddusb %%mm1, %%mm3 \n\t" +// "paddusb "MANGLE(b01)", %%mm3 \n\t" + + "movq (%%"REG_a", %1, 2), %%mm6 \n\t" //l3 + "movq (%0, %1, 4), %%mm5 \n\t" //l4 + "movq (%0, %1, 4), %%mm4 \n\t" //l4 + "psubusb %%mm6, %%mm5 \n\t" + "psubusb %%mm4, %%mm6 \n\t" + "por %%mm6, %%mm5 \n\t" // |l3-l4| + "pcmpeqb %%mm7, %%mm6 \n\t" // SIGN(l3-l4) + "pxor %%mm6, %%mm0 \n\t" + "pand %%mm0, %%mm3 \n\t" + PMINUB(%%mm5, %%mm3, %%mm0) + + "psubusb "MANGLE(b01)", %%mm3 \n\t" + PAVGB(%%mm7, %%mm3) + + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" + "movq (%0, %1, 4), %%mm2 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "psubb %%mm3, %%mm0 \n\t" + "paddb %%mm3, %%mm2 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "movq %%mm0, (%%"REG_a", %1, 2) \n\t" + "movq %%mm2, (%0, %1, 4) \n\t" #endif //0 - "lea (%0, %1), %%"REG_a" \n\t" - "pcmpeqb %%mm6, %%mm6 \n\t" // -1 -// 0 1 2 3 4 5 6 7 -// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 ecx+%1 ecx+2%1 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 - - - "movq (%%"REG_a", %1, 2), %%mm1 \n\t" // l3 - "movq (%0, %1, 4), %%mm0 \n\t" // l4 - "pxor %%mm6, %%mm1 \n\t" // -l3-1 - PAVGB(%%mm1, %%mm0) // -q+128 = (l4-l3+256)/2 + "lea (%0, %1), %%"REG_a" \n\t" + "pcmpeqb %%mm6, %%mm6 \n\t" // -1 +// 0 1 2 3 4 5 6 7 +// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 ecx+%1 ecx+2%1 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 + + + "movq (%%"REG_a", %1, 2), %%mm1 \n\t" // l3 + "movq (%0, %1, 4), %%mm0 \n\t" // l4 + "pxor %%mm6, %%mm1 \n\t" // -l3-1 + PAVGB(%%mm1, %%mm0) // -q+128 = (l4-l3+256)/2 // mm1=-l3-1, mm0=128-q - "movq (%%"REG_a", %1, 4), %%mm2 \n\t" // l5 - "movq (%%"REG_a", %1), %%mm3 \n\t" // l2 - "pxor %%mm6, %%mm2 \n\t" // -l5-1 - "movq %%mm2, %%mm5 \n\t" // -l5-1 - "movq "MANGLE(b80)", %%mm4 \n\t" // 128 - "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" - PAVGB(%%mm3, %%mm2) // (l2-l5+256)/2 - PAVGB(%%mm0, %%mm4) // ~(l4-l3)/4 + 128 - PAVGB(%%mm2, %%mm4) // ~(l2-l5)/4 +(l4-l3)/8 + 128 - PAVGB(%%mm0, %%mm4) // ~(l2-l5)/8 +5(l4-l3)/16 + 128 + "movq (%%"REG_a", %1, 4), %%mm2 \n\t" // l5 + "movq (%%"REG_a", %1), %%mm3 \n\t" // l2 + "pxor %%mm6, %%mm2 \n\t" // -l5-1 + "movq %%mm2, %%mm5 \n\t" // -l5-1 + "movq "MANGLE(b80)", %%mm4 \n\t" // 128 + "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" + PAVGB(%%mm3, %%mm2) // (l2-l5+256)/2 + PAVGB(%%mm0, %%mm4) // ~(l4-l3)/4 + 128 + PAVGB(%%mm2, %%mm4) // ~(l2-l5)/4 +(l4-l3)/8 + 128 + PAVGB(%%mm0, %%mm4) // ~(l2-l5)/8 +5(l4-l3)/16 + 128 // mm1=-l3-1, mm0=128-q, mm3=l2, mm4=menergy/16 + 128, mm5= -l5-1 - "movq (%%"REG_a"), %%mm2 \n\t" // l1 - "pxor %%mm6, %%mm2 \n\t" // -l1-1 - PAVGB(%%mm3, %%mm2) // (l2-l1+256)/2 - PAVGB((%0), %%mm1) // (l0-l3+256)/2 - "movq "MANGLE(b80)", %%mm3 \n\t" // 128 - PAVGB(%%mm2, %%mm3) // ~(l2-l1)/4 + 128 - PAVGB(%%mm1, %%mm3) // ~(l0-l3)/4 +(l2-l1)/8 + 128 - PAVGB(%%mm2, %%mm3) // ~(l0-l3)/8 +5(l2-l1)/16 + 128 + "movq (%%"REG_a"), %%mm2 \n\t" // l1 + "pxor %%mm6, %%mm2 \n\t" // -l1-1 + PAVGB(%%mm3, %%mm2) // (l2-l1+256)/2 + PAVGB((%0), %%mm1) // (l0-l3+256)/2 + "movq "MANGLE(b80)", %%mm3 \n\t" // 128 + PAVGB(%%mm2, %%mm3) // ~(l2-l1)/4 + 128 + PAVGB(%%mm1, %%mm3) // ~(l0-l3)/4 +(l2-l1)/8 + 128 + PAVGB(%%mm2, %%mm3) // ~(l0-l3)/8 +5(l2-l1)/16 + 128 // mm0=128-q, mm3=lenergy/16 + 128, mm4= menergy/16 + 128, mm5= -l5-1 - PAVGB((%%REGc, %1), %%mm5) // (l6-l5+256)/2 - "movq (%%"REG_c", %1, 2), %%mm1 \n\t" // l7 - "pxor %%mm6, %%mm1 \n\t" // -l7-1 - PAVGB((%0, %1, 4), %%mm1) // (l4-l7+256)/2 - "movq "MANGLE(b80)", %%mm2 \n\t" // 128 - PAVGB(%%mm5, %%mm2) // ~(l6-l5)/4 + 128 - PAVGB(%%mm1, %%mm2) // ~(l4-l7)/4 +(l6-l5)/8 + 128 - PAVGB(%%mm5, %%mm2) // ~(l4-l7)/8 +5(l6-l5)/16 + 128 + PAVGB((%%REGc, %1), %%mm5) // (l6-l5+256)/2 + "movq (%%"REG_c", %1, 2), %%mm1 \n\t" // l7 + "pxor %%mm6, %%mm1 \n\t" // -l7-1 + PAVGB((%0, %1, 4), %%mm1) // (l4-l7+256)/2 + "movq "MANGLE(b80)", %%mm2 \n\t" // 128 + PAVGB(%%mm5, %%mm2) // ~(l6-l5)/4 + 128 + PAVGB(%%mm1, %%mm2) // ~(l4-l7)/4 +(l6-l5)/8 + 128 + PAVGB(%%mm5, %%mm2) // ~(l4-l7)/8 +5(l6-l5)/16 + 128 // mm0=128-q, mm2=renergy/16 + 128, mm3=lenergy/16 + 128, mm4= menergy/16 + 128 - "movq "MANGLE(b00)", %%mm1 \n\t" // 0 - "movq "MANGLE(b00)", %%mm5 \n\t" // 0 - "psubb %%mm2, %%mm1 \n\t" // 128 - renergy/16 - "psubb %%mm3, %%mm5 \n\t" // 128 - lenergy/16 - PMAXUB(%%mm1, %%mm2) // 128 + |renergy/16| - PMAXUB(%%mm5, %%mm3) // 128 + |lenergy/16| - PMINUB(%%mm2, %%mm3, %%mm1) // 128 + MIN(|lenergy|,|renergy|)/16 + "movq "MANGLE(b00)", %%mm1 \n\t" // 0 + "movq "MANGLE(b00)", %%mm5 \n\t" // 0 + "psubb %%mm2, %%mm1 \n\t" // 128 - renergy/16 + "psubb %%mm3, %%mm5 \n\t" // 128 - lenergy/16 + PMAXUB(%%mm1, %%mm2) // 128 + |renergy/16| + PMAXUB(%%mm5, %%mm3) // 128 + |lenergy/16| + PMINUB(%%mm2, %%mm3, %%mm1) // 128 + MIN(|lenergy|,|renergy|)/16 // mm0=128-q, mm3=128 + MIN(|lenergy|,|renergy|)/16, mm4= menergy/16 + 128 - "movq "MANGLE(b00)", %%mm7 \n\t" // 0 - "movq %2, %%mm2 \n\t" // QP - PAVGB(%%mm6, %%mm2) // 128 + QP/2 - "psubb %%mm6, %%mm2 \n\t" - - "movq %%mm4, %%mm1 \n\t" - "pcmpgtb %%mm7, %%mm1 \n\t" // SIGN(menergy) - "pxor %%mm1, %%mm4 \n\t" - "psubb %%mm1, %%mm4 \n\t" // 128 + |menergy|/16 - "pcmpgtb %%mm4, %%mm2 \n\t" // |menergy|/16 < QP/2 - "psubusb %%mm3, %%mm4 \n\t" //d=|menergy|/16 - MIN(|lenergy|,|renergy|)/16 + "movq "MANGLE(b00)", %%mm7 \n\t" // 0 + "movq %2, %%mm2 \n\t" // QP + PAVGB(%%mm6, %%mm2) // 128 + QP/2 + "psubb %%mm6, %%mm2 \n\t" + + "movq %%mm4, %%mm1 \n\t" + "pcmpgtb %%mm7, %%mm1 \n\t" // SIGN(menergy) + "pxor %%mm1, %%mm4 \n\t" + "psubb %%mm1, %%mm4 \n\t" // 128 + |menergy|/16 + "pcmpgtb %%mm4, %%mm2 \n\t" // |menergy|/16 < QP/2 + "psubusb %%mm3, %%mm4 \n\t" //d=|menergy|/16 - MIN(|lenergy|,|renergy|)/16 // mm0=128-q, mm1= SIGN(menergy), mm2= |menergy|/16 < QP/2, mm4= d/16 - "movq %%mm4, %%mm3 \n\t" // d - "psubusb "MANGLE(b01)", %%mm4 \n\t" - PAVGB(%%mm7, %%mm4) // d/32 - PAVGB(%%mm7, %%mm4) // (d + 32)/64 - "paddb %%mm3, %%mm4 \n\t" // 5d/64 - "pand %%mm2, %%mm4 \n\t" - - "movq "MANGLE(b80)", %%mm5 \n\t" // 128 - "psubb %%mm0, %%mm5 \n\t" // q - "paddsb %%mm6, %%mm5 \n\t" // fix bad rounding - "pcmpgtb %%mm5, %%mm7 \n\t" // SIGN(q) - "pxor %%mm7, %%mm5 \n\t" - - PMINUB(%%mm5, %%mm4, %%mm3) // MIN(|q|, 5d/64) - "pxor %%mm1, %%mm7 \n\t" // SIGN(d*q) - - "pand %%mm7, %%mm4 \n\t" - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" - "movq (%0, %1, 4), %%mm2 \n\t" - "pxor %%mm1, %%mm0 \n\t" - "pxor %%mm1, %%mm2 \n\t" - "paddb %%mm4, %%mm0 \n\t" - "psubb %%mm4, %%mm2 \n\t" - "pxor %%mm1, %%mm0 \n\t" - "pxor %%mm1, %%mm2 \n\t" - "movq %%mm0, (%%"REG_a", %1, 2) \n\t" - "movq %%mm2, (%0, %1, 4) \n\t" - - : - : "r" (src), "r" ((long)stride), "m" (c->pQPb) - : "%"REG_a, "%"REG_c - ); + "movq %%mm4, %%mm3 \n\t" // d + "psubusb "MANGLE(b01)", %%mm4 \n\t" + PAVGB(%%mm7, %%mm4) // d/32 + PAVGB(%%mm7, %%mm4) // (d + 32)/64 + "paddb %%mm3, %%mm4 \n\t" // 5d/64 + "pand %%mm2, %%mm4 \n\t" + + "movq "MANGLE(b80)", %%mm5 \n\t" // 128 + "psubb %%mm0, %%mm5 \n\t" // q + "paddsb %%mm6, %%mm5 \n\t" // fix bad rounding + "pcmpgtb %%mm5, %%mm7 \n\t" // SIGN(q) + "pxor %%mm7, %%mm5 \n\t" + + PMINUB(%%mm5, %%mm4, %%mm3) // MIN(|q|, 5d/64) + "pxor %%mm1, %%mm7 \n\t" // SIGN(d*q) + + "pand %%mm7, %%mm4 \n\t" + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" + "movq (%0, %1, 4), %%mm2 \n\t" + "pxor %%mm1, %%mm0 \n\t" + "pxor %%mm1, %%mm2 \n\t" + "paddb %%mm4, %%mm0 \n\t" + "psubb %%mm4, %%mm2 \n\t" + "pxor %%mm1, %%mm0 \n\t" + "pxor %%mm1, %%mm2 \n\t" + "movq %%mm0, (%%"REG_a", %1, 2) \n\t" + "movq %%mm2, (%0, %1, 4) \n\t" + + : + : "r" (src), "r" ((long)stride), "m" (c->pQPb) + : "%"REG_a, "%"REG_c + ); /* - { - int x; - src-= stride; - for(x=0; x<BLOCK_SIZE; x++) - { - const int middleEnergy= 5*(src[l5] - src[l4]) + 2*(src[l3] - src[l6]); - if(ABS(middleEnergy)< 8*QP) - { - const int q=(src[l4] - src[l5])/2; - const int leftEnergy= 5*(src[l3] - src[l2]) + 2*(src[l1] - src[l4]); - const int rightEnergy= 5*(src[l7] - src[l6]) + 2*(src[l5] - src[l8]); - - int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); - d= MAX(d, 0); - - d= (5*d + 32) >> 6; - d*= SIGN(-middleEnergy); - - if(q>0) - { - d= d<0 ? 0 : d; - d= d>q ? q : d; - } - else - { - d= d>0 ? 0 : d; - d= d<q ? q : d; - } - - src[l4]-= d; - src[l5]+= d; - } - src++; - } + { + int x; + src-= stride; + for(x=0; x<BLOCK_SIZE; x++) + { + const int middleEnergy= 5*(src[l5] - src[l4]) + 2*(src[l3] - src[l6]); + if(ABS(middleEnergy)< 8*QP) + { + const int q=(src[l4] - src[l5])/2; + const int leftEnergy= 5*(src[l3] - src[l2]) + 2*(src[l1] - src[l4]); + const int rightEnergy= 5*(src[l7] - src[l6]) + 2*(src[l5] - src[l8]); + + int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); + d= MAX(d, 0); + + d= (5*d + 32) >> 6; + d*= SIGN(-middleEnergy); + + if(q>0) + { + d= d<0 ? 0 : d; + d= d>q ? q : d; + } + else + { + d= d>0 ? 0 : d; + d= d<q ? q : d; + } + + src[l4]-= d; + src[l5]+= d; + } + src++; + } src-=8; - for(x=0; x<8; x++) - { - int y; - for(y=4; y<6; y++) - { - int d= src[x+y*stride] - tmp[x+(y-4)*8]; - int ad= ABS(d); - static int max=0; - static int sum=0; - static int num=0; - static int bias=0; - - if(max<ad) max=ad; - sum+= ad>3 ? 1 : 0; - if(ad>3) - { - src[0] = src[7] = src[stride*7] = src[(stride+1)*7]=255; - } - if(y==4) bias+=d; - num++; - if(num%1000000 == 0) - { - printf(" %d %d %d %d\n", num, sum, max, bias); - } - } - } + for(x=0; x<8; x++) + { + int y; + for(y=4; y<6; y++) + { + int d= src[x+y*stride] - tmp[x+(y-4)*8]; + int ad= ABS(d); + static int max=0; + static int sum=0; + static int num=0; + static int bias=0; + + if(max<ad) max=ad; + sum+= ad>3 ? 1 : 0; + if(ad>3) + { + src[0] = src[7] = src[stride*7] = src[(stride+1)*7]=255; + } + if(y==4) bias+=d; + num++; + if(num%1000000 == 0) + { + printf(" %d %d %d %d\n", num, sum, max, bias); + } + } + } } */ #elif defined (HAVE_MMX) - src+= stride*4; - asm volatile( - "pxor %%mm7, %%mm7 \n\t" - "lea -40(%%"REG_SP"), %%"REG_c" \n\t" // make space for 4 8-byte vars - "and "ALIGN_MASK", %%"REG_c" \n\t" // align -// 0 1 2 3 4 5 6 7 -// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 edx+%1 edx+2%1 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 - - "movq (%0), %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" // low part of line 0 - "punpckhbw %%mm7, %%mm1 \n\t" // high part of line 0 - - "movq (%0, %1), %%mm2 \n\t" - "lea (%0, %1, 2), %%"REG_a" \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // low part of line 1 - "punpckhbw %%mm7, %%mm3 \n\t" // high part of line 1 - - "movq (%%"REG_a"), %%mm4 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" // low part of line 2 - "punpckhbw %%mm7, %%mm5 \n\t" // high part of line 2 - - "paddw %%mm0, %%mm0 \n\t" // 2L0 - "paddw %%mm1, %%mm1 \n\t" // 2H0 - "psubw %%mm4, %%mm2 \n\t" // L1 - L2 - "psubw %%mm5, %%mm3 \n\t" // H1 - H2 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - L1 + L2 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - H1 + H2 - - "psllw $2, %%mm2 \n\t" // 4L1 - 4L2 - "psllw $2, %%mm3 \n\t" // 4H1 - 4H2 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - - "movq (%%"REG_a", %1), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L3 - "punpckhbw %%mm7, %%mm3 \n\t" // H3 - - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - L3 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - H3 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 - "movq %%mm0, (%%"REG_c") \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "movq %%mm1, 8(%%"REG_c") \n\t" // 2H0 - 5H1 + 5H2 - 2H3 - - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" // L4 - "punpckhbw %%mm7, %%mm1 \n\t" // H4 - - "psubw %%mm0, %%mm2 \n\t" // L3 - L4 - "psubw %%mm1, %%mm3 \n\t" // H3 - H4 - "movq %%mm2, 16(%%"REG_c") \n\t" // L3 - L4 - "movq %%mm3, 24(%%"REG_c") \n\t" // H3 - H4 - "paddw %%mm4, %%mm4 \n\t" // 2L2 - "paddw %%mm5, %%mm5 \n\t" // 2H2 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - L3 + L4 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - H3 + H4 - - "lea (%%"REG_a", %1), %0 \n\t" - "psllw $2, %%mm2 \n\t" // 4L3 - 4L4 - "psllw $2, %%mm3 \n\t" // 4H3 - 4H4 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 + src+= stride*4; + asm volatile( + "pxor %%mm7, %%mm7 \n\t" + "lea -40(%%"REG_SP"), %%"REG_c" \n\t" // make space for 4 8-byte vars + "and "ALIGN_MASK", %%"REG_c" \n\t" // align +// 0 1 2 3 4 5 6 7 +// %0 %0+%1 %0+2%1 eax+2%1 %0+4%1 eax+4%1 edx+%1 edx+2%1 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 + + "movq (%0), %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" // low part of line 0 + "punpckhbw %%mm7, %%mm1 \n\t" // high part of line 0 + + "movq (%0, %1), %%mm2 \n\t" + "lea (%0, %1, 2), %%"REG_a" \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // low part of line 1 + "punpckhbw %%mm7, %%mm3 \n\t" // high part of line 1 + + "movq (%%"REG_a"), %%mm4 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" // low part of line 2 + "punpckhbw %%mm7, %%mm5 \n\t" // high part of line 2 + + "paddw %%mm0, %%mm0 \n\t" // 2L0 + "paddw %%mm1, %%mm1 \n\t" // 2H0 + "psubw %%mm4, %%mm2 \n\t" // L1 - L2 + "psubw %%mm5, %%mm3 \n\t" // H1 - H2 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - L1 + L2 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - H1 + H2 + + "psllw $2, %%mm2 \n\t" // 4L1 - 4L2 + "psllw $2, %%mm3 \n\t" // 4H1 - 4H2 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 + + "movq (%%"REG_a", %1), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L3 + "punpckhbw %%mm7, %%mm3 \n\t" // H3 + + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - L3 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - H3 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + "movq %%mm0, (%%"REG_c") \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "movq %%mm1, 8(%%"REG_c") \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" // L4 + "punpckhbw %%mm7, %%mm1 \n\t" // H4 + + "psubw %%mm0, %%mm2 \n\t" // L3 - L4 + "psubw %%mm1, %%mm3 \n\t" // H3 - H4 + "movq %%mm2, 16(%%"REG_c") \n\t" // L3 - L4 + "movq %%mm3, 24(%%"REG_c") \n\t" // H3 - H4 + "paddw %%mm4, %%mm4 \n\t" // 2L2 + "paddw %%mm5, %%mm5 \n\t" // 2H2 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - L3 + L4 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - H3 + H4 + + "lea (%%"REG_a", %1), %0 \n\t" + "psllw $2, %%mm2 \n\t" // 4L3 - 4L4 + "psllw $2, %%mm3 \n\t" // 4H3 - 4H4 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 //50 opcodes so far - "movq (%0, %1, 2), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L5 - "punpckhbw %%mm7, %%mm3 \n\t" // H5 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - L5 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - H5 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - 2L5 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - 2H5 - - "movq (%%"REG_a", %1, 4), %%mm6 \n\t" - "punpcklbw %%mm7, %%mm6 \n\t" // L6 - "psubw %%mm6, %%mm2 \n\t" // L5 - L6 - "movq (%%"REG_a", %1, 4), %%mm6 \n\t" - "punpckhbw %%mm7, %%mm6 \n\t" // H6 - "psubw %%mm6, %%mm3 \n\t" // H5 - H6 - - "paddw %%mm0, %%mm0 \n\t" // 2L4 - "paddw %%mm1, %%mm1 \n\t" // 2H4 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - L5 + L6 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - H5 + H6 - - "psllw $2, %%mm2 \n\t" // 4L5 - 4L6 - "psllw $2, %%mm3 \n\t" // 4H5 - 4H6 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - - "movq (%0, %1, 4), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L7 - "punpckhbw %%mm7, %%mm3 \n\t" // H7 - - "paddw %%mm2, %%mm2 \n\t" // 2L7 - "paddw %%mm3, %%mm3 \n\t" // 2H7 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - 2L7 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - 2H7 - - "movq (%%"REG_c"), %%mm2 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "movq 8(%%"REG_c"), %%mm3 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + "movq (%0, %1, 2), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L5 + "punpckhbw %%mm7, %%mm3 \n\t" // H5 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - L5 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - H5 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - 2L5 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - 2H5 + + "movq (%%"REG_a", %1, 4), %%mm6 \n\t" + "punpcklbw %%mm7, %%mm6 \n\t" // L6 + "psubw %%mm6, %%mm2 \n\t" // L5 - L6 + "movq (%%"REG_a", %1, 4), %%mm6 \n\t" + "punpckhbw %%mm7, %%mm6 \n\t" // H6 + "psubw %%mm6, %%mm3 \n\t" // H5 - H6 + + "paddw %%mm0, %%mm0 \n\t" // 2L4 + "paddw %%mm1, %%mm1 \n\t" // 2H4 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - L5 + L6 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - H5 + H6 + + "psllw $2, %%mm2 \n\t" // 4L5 - 4L6 + "psllw $2, %%mm3 \n\t" // 4H5 - 4H6 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 + + "movq (%0, %1, 4), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L7 + "punpckhbw %%mm7, %%mm3 \n\t" // H7 + + "paddw %%mm2, %%mm2 \n\t" // 2L7 + "paddw %%mm3, %%mm3 \n\t" // 2H7 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - 2L7 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - 2H7 + + "movq (%%"REG_c"), %%mm2 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "movq 8(%%"REG_c"), %%mm3 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 #ifdef HAVE_MMX2 - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm0, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm1, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm2, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm3, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm0, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm1, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm2, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm3, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| #else - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm0, %%mm6 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "psubw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm1, %%mm6 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "psubw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm2, %%mm6 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "psubw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm3, %%mm6 \n\t" - "pxor %%mm6, %%mm3 \n\t" - "psubw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm0, %%mm6 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "psubw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm1, %%mm6 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "psubw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm2, %%mm6 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "psubw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm3, %%mm6 \n\t" + "pxor %%mm6, %%mm3 \n\t" + "psubw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| #endif #ifdef HAVE_MMX2 - "pminsw %%mm2, %%mm0 \n\t" - "pminsw %%mm3, %%mm1 \n\t" + "pminsw %%mm2, %%mm0 \n\t" + "pminsw %%mm3, %%mm1 \n\t" #else - "movq %%mm0, %%mm6 \n\t" - "psubusw %%mm2, %%mm6 \n\t" - "psubw %%mm6, %%mm0 \n\t" - "movq %%mm1, %%mm6 \n\t" - "psubusw %%mm3, %%mm6 \n\t" - "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, %%mm6 \n\t" + "psubusw %%mm2, %%mm6 \n\t" + "psubw %%mm6, %%mm0 \n\t" + "movq %%mm1, %%mm6 \n\t" + "psubusw %%mm3, %%mm6 \n\t" + "psubw %%mm6, %%mm1 \n\t" #endif - "movd %2, %%mm2 \n\t" // QP - "punpcklbw %%mm7, %%mm2 \n\t" - - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm4, %%mm6 \n\t" // sign(2L2 - 5L3 + 5L4 - 2L5) - "pxor %%mm6, %%mm4 \n\t" - "psubw %%mm6, %%mm4 \n\t" // |2L2 - 5L3 + 5L4 - 2L5| - "pcmpgtw %%mm5, %%mm7 \n\t" // sign(2H2 - 5H3 + 5H4 - 2H5) - "pxor %%mm7, %%mm5 \n\t" - "psubw %%mm7, %%mm5 \n\t" // |2H2 - 5H3 + 5H4 - 2H5| + "movd %2, %%mm2 \n\t" // QP + "punpcklbw %%mm7, %%mm2 \n\t" + + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm4, %%mm6 \n\t" // sign(2L2 - 5L3 + 5L4 - 2L5) + "pxor %%mm6, %%mm4 \n\t" + "psubw %%mm6, %%mm4 \n\t" // |2L2 - 5L3 + 5L4 - 2L5| + "pcmpgtw %%mm5, %%mm7 \n\t" // sign(2H2 - 5H3 + 5H4 - 2H5) + "pxor %%mm7, %%mm5 \n\t" + "psubw %%mm7, %%mm5 \n\t" // |2H2 - 5H3 + 5H4 - 2H5| // 100 opcodes - "psllw $3, %%mm2 \n\t" // 8QP - "movq %%mm2, %%mm3 \n\t" // 8QP - "pcmpgtw %%mm4, %%mm2 \n\t" - "pcmpgtw %%mm5, %%mm3 \n\t" - "pand %%mm2, %%mm4 \n\t" - "pand %%mm3, %%mm5 \n\t" - - - "psubusw %%mm0, %%mm4 \n\t" // hd - "psubusw %%mm1, %%mm5 \n\t" // ld - - - "movq "MANGLE(w05)", %%mm2 \n\t" // 5 - "pmullw %%mm2, %%mm4 \n\t" - "pmullw %%mm2, %%mm5 \n\t" - "movq "MANGLE(w20)", %%mm2 \n\t" // 32 - "paddw %%mm2, %%mm4 \n\t" - "paddw %%mm2, %%mm5 \n\t" - "psrlw $6, %%mm4 \n\t" - "psrlw $6, %%mm5 \n\t" - - "movq 16(%%"REG_c"), %%mm0 \n\t" // L3 - L4 - "movq 24(%%"REG_c"), %%mm1 \n\t" // H3 - H4 - - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - - "pcmpgtw %%mm0, %%mm2 \n\t" // sign (L3-L4) - "pcmpgtw %%mm1, %%mm3 \n\t" // sign (H3-H4) - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // |L3-L4| - "psubw %%mm3, %%mm1 \n\t" // |H3-H4| - "psrlw $1, %%mm0 \n\t" // |L3 - L4|/2 - "psrlw $1, %%mm1 \n\t" // |H3 - H4|/2 - - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm7, %%mm3 \n\t" - "pand %%mm2, %%mm4 \n\t" - "pand %%mm3, %%mm5 \n\t" + "psllw $3, %%mm2 \n\t" // 8QP + "movq %%mm2, %%mm3 \n\t" // 8QP + "pcmpgtw %%mm4, %%mm2 \n\t" + "pcmpgtw %%mm5, %%mm3 \n\t" + "pand %%mm2, %%mm4 \n\t" + "pand %%mm3, %%mm5 \n\t" + + + "psubusw %%mm0, %%mm4 \n\t" // hd + "psubusw %%mm1, %%mm5 \n\t" // ld + + + "movq "MANGLE(w05)", %%mm2 \n\t" // 5 + "pmullw %%mm2, %%mm4 \n\t" + "pmullw %%mm2, %%mm5 \n\t" + "movq "MANGLE(w20)", %%mm2 \n\t" // 32 + "paddw %%mm2, %%mm4 \n\t" + "paddw %%mm2, %%mm5 \n\t" + "psrlw $6, %%mm4 \n\t" + "psrlw $6, %%mm5 \n\t" + + "movq 16(%%"REG_c"), %%mm0 \n\t" // L3 - L4 + "movq 24(%%"REG_c"), %%mm1 \n\t" // H3 - H4 + + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + + "pcmpgtw %%mm0, %%mm2 \n\t" // sign (L3-L4) + "pcmpgtw %%mm1, %%mm3 \n\t" // sign (H3-H4) + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // |L3-L4| + "psubw %%mm3, %%mm1 \n\t" // |H3-H4| + "psrlw $1, %%mm0 \n\t" // |L3 - L4|/2 + "psrlw $1, %%mm1 \n\t" // |H3 - H4|/2 + + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm7, %%mm3 \n\t" + "pand %%mm2, %%mm4 \n\t" + "pand %%mm3, %%mm5 \n\t" #ifdef HAVE_MMX2 - "pminsw %%mm0, %%mm4 \n\t" - "pminsw %%mm1, %%mm5 \n\t" + "pminsw %%mm0, %%mm4 \n\t" + "pminsw %%mm1, %%mm5 \n\t" #else - "movq %%mm4, %%mm2 \n\t" - "psubusw %%mm0, %%mm2 \n\t" - "psubw %%mm2, %%mm4 \n\t" - "movq %%mm5, %%mm2 \n\t" - "psubusw %%mm1, %%mm2 \n\t" - "psubw %%mm2, %%mm5 \n\t" + "movq %%mm4, %%mm2 \n\t" + "psubusw %%mm0, %%mm2 \n\t" + "psubw %%mm2, %%mm4 \n\t" + "movq %%mm5, %%mm2 \n\t" + "psubusw %%mm1, %%mm2 \n\t" + "psubw %%mm2, %%mm5 \n\t" #endif - "pxor %%mm6, %%mm4 \n\t" - "pxor %%mm7, %%mm5 \n\t" - "psubw %%mm6, %%mm4 \n\t" - "psubw %%mm7, %%mm5 \n\t" - "packsswb %%mm5, %%mm4 \n\t" - "movq (%0), %%mm0 \n\t" - "paddb %%mm4, %%mm0 \n\t" - "movq %%mm0, (%0) \n\t" - "movq (%0, %1), %%mm0 \n\t" - "psubb %%mm4, %%mm0 \n\t" - "movq %%mm0, (%0, %1) \n\t" - - : "+r" (src) - : "r" ((long)stride), "m" (c->pQPb) - : "%"REG_a, "%"REG_c - ); + "pxor %%mm6, %%mm4 \n\t" + "pxor %%mm7, %%mm5 \n\t" + "psubw %%mm6, %%mm4 \n\t" + "psubw %%mm7, %%mm5 \n\t" + "packsswb %%mm5, %%mm4 \n\t" + "movq (%0), %%mm0 \n\t" + "paddb %%mm4, %%mm0 \n\t" + "movq %%mm0, (%0) \n\t" + "movq (%0, %1), %%mm0 \n\t" + "psubb %%mm4, %%mm0 \n\t" + "movq %%mm0, (%0, %1) \n\t" + + : "+r" (src) + : "r" ((long)stride), "m" (c->pQPb) + : "%"REG_a, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - const int l1= stride; - const int l2= stride + l1; - const int l3= stride + l2; - const int l4= stride + l3; - const int l5= stride + l4; - const int l6= stride + l5; - const int l7= stride + l6; - const int l8= stride + l7; -// const int l9= stride + l8; - int x; - src+= stride*3; - for(x=0; x<BLOCK_SIZE; x++) - { - const int middleEnergy= 5*(src[l5] - src[l4]) + 2*(src[l3] - src[l6]); - if(ABS(middleEnergy) < 8*c->QP) - { - const int q=(src[l4] - src[l5])/2; - const int leftEnergy= 5*(src[l3] - src[l2]) + 2*(src[l1] - src[l4]); - const int rightEnergy= 5*(src[l7] - src[l6]) + 2*(src[l5] - src[l8]); - - int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); - d= MAX(d, 0); - - d= (5*d + 32) >> 6; - d*= SIGN(-middleEnergy); - - if(q>0) - { - d= d<0 ? 0 : d; - d= d>q ? q : d; - } - else - { - d= d>0 ? 0 : d; - d= d<q ? q : d; - } - - src[l4]-= d; - src[l5]+= d; - } - src++; - } + const int l1= stride; + const int l2= stride + l1; + const int l3= stride + l2; + const int l4= stride + l3; + const int l5= stride + l4; + const int l6= stride + l5; + const int l7= stride + l6; + const int l8= stride + l7; +// const int l9= stride + l8; + int x; + src+= stride*3; + for(x=0; x<BLOCK_SIZE; x++) + { + const int middleEnergy= 5*(src[l5] - src[l4]) + 2*(src[l3] - src[l6]); + if(ABS(middleEnergy) < 8*c->QP) + { + const int q=(src[l4] - src[l5])/2; + const int leftEnergy= 5*(src[l3] - src[l2]) + 2*(src[l1] - src[l4]); + const int rightEnergy= 5*(src[l7] - src[l6]) + 2*(src[l5] - src[l8]); + + int d= ABS(middleEnergy) - MIN( ABS(leftEnergy), ABS(rightEnergy) ); + d= MAX(d, 0); + + d= (5*d + 32) >> 6; + d*= SIGN(-middleEnergy); + + if(q>0) + { + d= d<0 ? 0 : d; + d= d>q ? q : d; + } + else + { + d= d>0 ? 0 : d; + d= d<q ? q : d; + } + + src[l4]-= d; + src[l5]+= d; + } + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } #endif //HAVE_ALTIVEC @@ -1183,36 +1183,36 @@ static inline void RENAME(dering)(uint8_t src[], int stride, PPContext *c) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - asm volatile( - "pxor %%mm6, %%mm6 \n\t" - "pcmpeqb %%mm7, %%mm7 \n\t" - "movq %2, %%mm0 \n\t" - "punpcklbw %%mm6, %%mm0 \n\t" - "psrlw $1, %%mm0 \n\t" - "psubw %%mm7, %%mm0 \n\t" - "packuswb %%mm0, %%mm0 \n\t" - "movq %%mm0, %3 \n\t" - - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" - -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + asm volatile( + "pxor %%mm6, %%mm6 \n\t" + "pcmpeqb %%mm7, %%mm7 \n\t" + "movq %2, %%mm0 \n\t" + "punpcklbw %%mm6, %%mm0 \n\t" + "psrlw $1, %%mm0 \n\t" + "psubw %%mm7, %%mm0 \n\t" + "packuswb %%mm0, %%mm0 \n\t" + "movq %%mm0, %3 \n\t" + + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" + +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 #undef FIND_MIN_MAX #ifdef HAVE_MMX2 #define REAL_FIND_MIN_MAX(addr)\ - "movq " #addr ", %%mm0 \n\t"\ - "pminub %%mm0, %%mm7 \n\t"\ - "pmaxub %%mm0, %%mm6 \n\t" + "movq " #addr ", %%mm0 \n\t"\ + "pminub %%mm0, %%mm7 \n\t"\ + "pmaxub %%mm0, %%mm6 \n\t" #else #define REAL_FIND_MIN_MAX(addr)\ - "movq " #addr ", %%mm0 \n\t"\ - "movq %%mm7, %%mm1 \n\t"\ - "psubusb %%mm0, %%mm6 \n\t"\ - "paddb %%mm0, %%mm6 \n\t"\ - "psubusb %%mm0, %%mm1 \n\t"\ - "psubb %%mm1, %%mm7 \n\t" + "movq " #addr ", %%mm0 \n\t"\ + "movq %%mm7, %%mm1 \n\t"\ + "psubusb %%mm0, %%mm6 \n\t"\ + "paddb %%mm0, %%mm6 \n\t"\ + "psubusb %%mm0, %%mm1 \n\t"\ + "psubb %%mm1, %%mm7 \n\t" #endif #define FIND_MIN_MAX(addr) REAL_FIND_MIN_MAX(addr) @@ -1225,155 +1225,155 @@ FIND_MIN_MAX((%%REGd, %1, 2)) FIND_MIN_MAX((%0, %1, 8)) - "movq %%mm7, %%mm4 \n\t" - "psrlq $8, %%mm7 \n\t" + "movq %%mm7, %%mm4 \n\t" + "psrlq $8, %%mm7 \n\t" #ifdef HAVE_MMX2 - "pminub %%mm4, %%mm7 \n\t" // min of pixels - "pshufw $0xF9, %%mm7, %%mm4 \n\t" - "pminub %%mm4, %%mm7 \n\t" // min of pixels - "pshufw $0xFE, %%mm7, %%mm4 \n\t" - "pminub %%mm4, %%mm7 \n\t" + "pminub %%mm4, %%mm7 \n\t" // min of pixels + "pshufw $0xF9, %%mm7, %%mm4 \n\t" + "pminub %%mm4, %%mm7 \n\t" // min of pixels + "pshufw $0xFE, %%mm7, %%mm4 \n\t" + "pminub %%mm4, %%mm7 \n\t" #else - "movq %%mm7, %%mm1 \n\t" - "psubusb %%mm4, %%mm1 \n\t" - "psubb %%mm1, %%mm7 \n\t" - "movq %%mm7, %%mm4 \n\t" - "psrlq $16, %%mm7 \n\t" - "movq %%mm7, %%mm1 \n\t" - "psubusb %%mm4, %%mm1 \n\t" - "psubb %%mm1, %%mm7 \n\t" - "movq %%mm7, %%mm4 \n\t" - "psrlq $32, %%mm7 \n\t" - "movq %%mm7, %%mm1 \n\t" - "psubusb %%mm4, %%mm1 \n\t" - "psubb %%mm1, %%mm7 \n\t" + "movq %%mm7, %%mm1 \n\t" + "psubusb %%mm4, %%mm1 \n\t" + "psubb %%mm1, %%mm7 \n\t" + "movq %%mm7, %%mm4 \n\t" + "psrlq $16, %%mm7 \n\t" + "movq %%mm7, %%mm1 \n\t" + "psubusb %%mm4, %%mm1 \n\t" + "psubb %%mm1, %%mm7 \n\t" + "movq %%mm7, %%mm4 \n\t" + "psrlq $32, %%mm7 \n\t" + "movq %%mm7, %%mm1 \n\t" + "psubusb %%mm4, %%mm1 \n\t" + "psubb %%mm1, %%mm7 \n\t" #endif - "movq %%mm6, %%mm4 \n\t" - "psrlq $8, %%mm6 \n\t" + "movq %%mm6, %%mm4 \n\t" + "psrlq $8, %%mm6 \n\t" #ifdef HAVE_MMX2 - "pmaxub %%mm4, %%mm6 \n\t" // max of pixels - "pshufw $0xF9, %%mm6, %%mm4 \n\t" - "pmaxub %%mm4, %%mm6 \n\t" - "pshufw $0xFE, %%mm6, %%mm4 \n\t" - "pmaxub %%mm4, %%mm6 \n\t" + "pmaxub %%mm4, %%mm6 \n\t" // max of pixels + "pshufw $0xF9, %%mm6, %%mm4 \n\t" + "pmaxub %%mm4, %%mm6 \n\t" + "pshufw $0xFE, %%mm6, %%mm4 \n\t" + "pmaxub %%mm4, %%mm6 \n\t" #else - "psubusb %%mm4, %%mm6 \n\t" - "paddb %%mm4, %%mm6 \n\t" - "movq %%mm6, %%mm4 \n\t" - "psrlq $16, %%mm6 \n\t" - "psubusb %%mm4, %%mm6 \n\t" - "paddb %%mm4, %%mm6 \n\t" - "movq %%mm6, %%mm4 \n\t" - "psrlq $32, %%mm6 \n\t" - "psubusb %%mm4, %%mm6 \n\t" - "paddb %%mm4, %%mm6 \n\t" + "psubusb %%mm4, %%mm6 \n\t" + "paddb %%mm4, %%mm6 \n\t" + "movq %%mm6, %%mm4 \n\t" + "psrlq $16, %%mm6 \n\t" + "psubusb %%mm4, %%mm6 \n\t" + "paddb %%mm4, %%mm6 \n\t" + "movq %%mm6, %%mm4 \n\t" + "psrlq $32, %%mm6 \n\t" + "psubusb %%mm4, %%mm6 \n\t" + "paddb %%mm4, %%mm6 \n\t" #endif - "movq %%mm6, %%mm0 \n\t" // max - "psubb %%mm7, %%mm6 \n\t" // max - min - "movd %%mm6, %%ecx \n\t" - "cmpb "MANGLE(deringThreshold)", %%cl \n\t" - " jb 1f \n\t" - "lea -24(%%"REG_SP"), %%"REG_c" \n\t" - "and "ALIGN_MASK", %%"REG_c" \n\t" - PAVGB(%%mm0, %%mm7) // a=(max + min)/2 - "punpcklbw %%mm7, %%mm7 \n\t" - "punpcklbw %%mm7, %%mm7 \n\t" - "punpcklbw %%mm7, %%mm7 \n\t" - "movq %%mm7, (%%"REG_c") \n\t" - - "movq (%0), %%mm0 \n\t" // L10 - "movq %%mm0, %%mm1 \n\t" // L10 - "movq %%mm0, %%mm2 \n\t" // L10 - "psllq $8, %%mm1 \n\t" - "psrlq $8, %%mm2 \n\t" - "movd -4(%0), %%mm3 \n\t" - "movd 8(%0), %%mm4 \n\t" - "psrlq $24, %%mm3 \n\t" - "psllq $56, %%mm4 \n\t" - "por %%mm3, %%mm1 \n\t" // L00 - "por %%mm4, %%mm2 \n\t" // L20 - "movq %%mm1, %%mm3 \n\t" // L00 - PAVGB(%%mm2, %%mm1) // (L20 + L00)/2 - PAVGB(%%mm0, %%mm1) // (L20 + L00 + 2L10)/4 - "psubusb %%mm7, %%mm0 \n\t" - "psubusb %%mm7, %%mm2 \n\t" - "psubusb %%mm7, %%mm3 \n\t" - "pcmpeqb "MANGLE(b00)", %%mm0 \n\t" // L10 > a ? 0 : -1 - "pcmpeqb "MANGLE(b00)", %%mm2 \n\t" // L20 > a ? 0 : -1 - "pcmpeqb "MANGLE(b00)", %%mm3 \n\t" // L00 > a ? 0 : -1 - "paddb %%mm2, %%mm0 \n\t" - "paddb %%mm3, %%mm0 \n\t" - - "movq (%%"REG_a"), %%mm2 \n\t" // L11 - "movq %%mm2, %%mm3 \n\t" // L11 - "movq %%mm2, %%mm4 \n\t" // L11 - "psllq $8, %%mm3 \n\t" - "psrlq $8, %%mm4 \n\t" - "movd -4(%%"REG_a"), %%mm5 \n\t" - "movd 8(%%"REG_a"), %%mm6 \n\t" - "psrlq $24, %%mm5 \n\t" - "psllq $56, %%mm6 \n\t" - "por %%mm5, %%mm3 \n\t" // L01 - "por %%mm6, %%mm4 \n\t" // L21 - "movq %%mm3, %%mm5 \n\t" // L01 - PAVGB(%%mm4, %%mm3) // (L21 + L01)/2 - PAVGB(%%mm2, %%mm3) // (L21 + L01 + 2L11)/4 - "psubusb %%mm7, %%mm2 \n\t" - "psubusb %%mm7, %%mm4 \n\t" - "psubusb %%mm7, %%mm5 \n\t" - "pcmpeqb "MANGLE(b00)", %%mm2 \n\t" // L11 > a ? 0 : -1 - "pcmpeqb "MANGLE(b00)", %%mm4 \n\t" // L21 > a ? 0 : -1 - "pcmpeqb "MANGLE(b00)", %%mm5 \n\t" // L01 > a ? 0 : -1 - "paddb %%mm4, %%mm2 \n\t" - "paddb %%mm5, %%mm2 \n\t" + "movq %%mm6, %%mm0 \n\t" // max + "psubb %%mm7, %%mm6 \n\t" // max - min + "movd %%mm6, %%ecx \n\t" + "cmpb "MANGLE(deringThreshold)", %%cl \n\t" + " jb 1f \n\t" + "lea -24(%%"REG_SP"), %%"REG_c" \n\t" + "and "ALIGN_MASK", %%"REG_c" \n\t" + PAVGB(%%mm0, %%mm7) // a=(max + min)/2 + "punpcklbw %%mm7, %%mm7 \n\t" + "punpcklbw %%mm7, %%mm7 \n\t" + "punpcklbw %%mm7, %%mm7 \n\t" + "movq %%mm7, (%%"REG_c") \n\t" + + "movq (%0), %%mm0 \n\t" // L10 + "movq %%mm0, %%mm1 \n\t" // L10 + "movq %%mm0, %%mm2 \n\t" // L10 + "psllq $8, %%mm1 \n\t" + "psrlq $8, %%mm2 \n\t" + "movd -4(%0), %%mm3 \n\t" + "movd 8(%0), %%mm4 \n\t" + "psrlq $24, %%mm3 \n\t" + "psllq $56, %%mm4 \n\t" + "por %%mm3, %%mm1 \n\t" // L00 + "por %%mm4, %%mm2 \n\t" // L20 + "movq %%mm1, %%mm3 \n\t" // L00 + PAVGB(%%mm2, %%mm1) // (L20 + L00)/2 + PAVGB(%%mm0, %%mm1) // (L20 + L00 + 2L10)/4 + "psubusb %%mm7, %%mm0 \n\t" + "psubusb %%mm7, %%mm2 \n\t" + "psubusb %%mm7, %%mm3 \n\t" + "pcmpeqb "MANGLE(b00)", %%mm0 \n\t" // L10 > a ? 0 : -1 + "pcmpeqb "MANGLE(b00)", %%mm2 \n\t" // L20 > a ? 0 : -1 + "pcmpeqb "MANGLE(b00)", %%mm3 \n\t" // L00 > a ? 0 : -1 + "paddb %%mm2, %%mm0 \n\t" + "paddb %%mm3, %%mm0 \n\t" + + "movq (%%"REG_a"), %%mm2 \n\t" // L11 + "movq %%mm2, %%mm3 \n\t" // L11 + "movq %%mm2, %%mm4 \n\t" // L11 + "psllq $8, %%mm3 \n\t" + "psrlq $8, %%mm4 \n\t" + "movd -4(%%"REG_a"), %%mm5 \n\t" + "movd 8(%%"REG_a"), %%mm6 \n\t" + "psrlq $24, %%mm5 \n\t" + "psllq $56, %%mm6 \n\t" + "por %%mm5, %%mm3 \n\t" // L01 + "por %%mm6, %%mm4 \n\t" // L21 + "movq %%mm3, %%mm5 \n\t" // L01 + PAVGB(%%mm4, %%mm3) // (L21 + L01)/2 + PAVGB(%%mm2, %%mm3) // (L21 + L01 + 2L11)/4 + "psubusb %%mm7, %%mm2 \n\t" + "psubusb %%mm7, %%mm4 \n\t" + "psubusb %%mm7, %%mm5 \n\t" + "pcmpeqb "MANGLE(b00)", %%mm2 \n\t" // L11 > a ? 0 : -1 + "pcmpeqb "MANGLE(b00)", %%mm4 \n\t" // L21 > a ? 0 : -1 + "pcmpeqb "MANGLE(b00)", %%mm5 \n\t" // L01 > a ? 0 : -1 + "paddb %%mm4, %%mm2 \n\t" + "paddb %%mm5, %%mm2 \n\t" // 0, 2, 3, 1 #define REAL_DERING_CORE(dst,src,ppsx,psx,sx,pplx,plx,lx,t0,t1) \ - "movq " #src ", " #sx " \n\t" /* src[0] */\ - "movq " #sx ", " #lx " \n\t" /* src[0] */\ - "movq " #sx ", " #t0 " \n\t" /* src[0] */\ - "psllq $8, " #lx " \n\t"\ - "psrlq $8, " #t0 " \n\t"\ - "movd -4" #src ", " #t1 " \n\t"\ - "psrlq $24, " #t1 " \n\t"\ - "por " #t1 ", " #lx " \n\t" /* src[-1] */\ - "movd 8" #src ", " #t1 " \n\t"\ - "psllq $56, " #t1 " \n\t"\ - "por " #t1 ", " #t0 " \n\t" /* src[+1] */\ - "movq " #lx ", " #t1 " \n\t" /* src[-1] */\ - PAVGB(t0, lx) /* (src[-1] + src[+1])/2 */\ - PAVGB(sx, lx) /* (src[-1] + 2src[0] + src[+1])/4 */\ - PAVGB(lx, pplx) \ - "movq " #lx ", 8(%%"REG_c") \n\t"\ - "movq (%%"REG_c"), " #lx " \n\t"\ - "psubusb " #lx ", " #t1 " \n\t"\ - "psubusb " #lx ", " #t0 " \n\t"\ - "psubusb " #lx ", " #sx " \n\t"\ - "movq "MANGLE(b00)", " #lx " \n\t"\ - "pcmpeqb " #lx ", " #t1 " \n\t" /* src[-1] > a ? 0 : -1*/\ - "pcmpeqb " #lx ", " #t0 " \n\t" /* src[+1] > a ? 0 : -1*/\ - "pcmpeqb " #lx ", " #sx " \n\t" /* src[0] > a ? 0 : -1*/\ - "paddb " #t1 ", " #t0 " \n\t"\ - "paddb " #t0 ", " #sx " \n\t"\ + "movq " #src ", " #sx " \n\t" /* src[0] */\ + "movq " #sx ", " #lx " \n\t" /* src[0] */\ + "movq " #sx ", " #t0 " \n\t" /* src[0] */\ + "psllq $8, " #lx " \n\t"\ + "psrlq $8, " #t0 " \n\t"\ + "movd -4" #src ", " #t1 " \n\t"\ + "psrlq $24, " #t1 " \n\t"\ + "por " #t1 ", " #lx " \n\t" /* src[-1] */\ + "movd 8" #src ", " #t1 " \n\t"\ + "psllq $56, " #t1 " \n\t"\ + "por " #t1 ", " #t0 " \n\t" /* src[+1] */\ + "movq " #lx ", " #t1 " \n\t" /* src[-1] */\ + PAVGB(t0, lx) /* (src[-1] + src[+1])/2 */\ + PAVGB(sx, lx) /* (src[-1] + 2src[0] + src[+1])/4 */\ + PAVGB(lx, pplx) \ + "movq " #lx ", 8(%%"REG_c") \n\t"\ + "movq (%%"REG_c"), " #lx " \n\t"\ + "psubusb " #lx ", " #t1 " \n\t"\ + "psubusb " #lx ", " #t0 " \n\t"\ + "psubusb " #lx ", " #sx " \n\t"\ + "movq "MANGLE(b00)", " #lx " \n\t"\ + "pcmpeqb " #lx ", " #t1 " \n\t" /* src[-1] > a ? 0 : -1*/\ + "pcmpeqb " #lx ", " #t0 " \n\t" /* src[+1] > a ? 0 : -1*/\ + "pcmpeqb " #lx ", " #sx " \n\t" /* src[0] > a ? 0 : -1*/\ + "paddb " #t1 ", " #t0 " \n\t"\ + "paddb " #t0 ", " #sx " \n\t"\ \ - PAVGB(plx, pplx) /* filtered */\ - "movq " #dst ", " #t0 " \n\t" /* dst */\ - "movq " #t0 ", " #t1 " \n\t" /* dst */\ - "psubusb %3, " #t0 " \n\t"\ - "paddusb %3, " #t1 " \n\t"\ - PMAXUB(t0, pplx)\ - PMINUB(t1, pplx, t0)\ - "paddb " #sx ", " #ppsx " \n\t"\ - "paddb " #psx ", " #ppsx " \n\t"\ - "#paddb "MANGLE(b02)", " #ppsx " \n\t"\ - "pand "MANGLE(b08)", " #ppsx " \n\t"\ - "pcmpeqb " #lx ", " #ppsx " \n\t"\ - "pand " #ppsx ", " #pplx " \n\t"\ - "pandn " #dst ", " #ppsx " \n\t"\ - "por " #pplx ", " #ppsx " \n\t"\ - "movq " #ppsx ", " #dst " \n\t"\ - "movq 8(%%"REG_c"), " #lx " \n\t" + PAVGB(plx, pplx) /* filtered */\ + "movq " #dst ", " #t0 " \n\t" /* dst */\ + "movq " #t0 ", " #t1 " \n\t" /* dst */\ + "psubusb %3, " #t0 " \n\t"\ + "paddusb %3, " #t1 " \n\t"\ + PMAXUB(t0, pplx)\ + PMINUB(t1, pplx, t0)\ + "paddb " #sx ", " #ppsx " \n\t"\ + "paddb " #psx ", " #ppsx " \n\t"\ + "#paddb "MANGLE(b02)", " #ppsx " \n\t"\ + "pand "MANGLE(b08)", " #ppsx " \n\t"\ + "pcmpeqb " #lx ", " #ppsx " \n\t"\ + "pand " #ppsx ", " #pplx " \n\t"\ + "pandn " #dst ", " #ppsx " \n\t"\ + "por " #pplx ", " #ppsx " \n\t"\ + "movq " #ppsx ", " #dst " \n\t"\ + "movq 8(%%"REG_c"), " #lx " \n\t" #define DERING_CORE(dst,src,ppsx,psx,sx,pplx,plx,lx,t0,t1) \ REAL_DERING_CORE(dst,src,ppsx,psx,sx,pplx,plx,lx,t0,t1) @@ -1392,149 +1392,149 @@ 1110111 */ -//DERING_CORE(dst,src ,ppsx ,psx ,sx ,pplx ,plx ,lx ,t0 ,t1) -DERING_CORE((%%REGa),(%%REGa, %1) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) -DERING_CORE((%%REGa, %1),(%%REGa, %1, 2) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) -DERING_CORE((%%REGa, %1, 2),(%0, %1, 4) ,%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7) -DERING_CORE((%0, %1, 4),(%%REGd) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) -DERING_CORE((%%REGd),(%%REGd, %1) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) -DERING_CORE((%%REGd, %1), (%%REGd, %1, 2),%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7) -DERING_CORE((%%REGd, %1, 2),(%0, %1, 8) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) -DERING_CORE((%0, %1, 8),(%%REGd, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) - - "1: \n\t" - : : "r" (src), "r" ((long)stride), "m" (c->pQPb), "m"(c->pQPb2) - : "%"REG_a, "%"REG_d, "%"REG_c - ); +//DERING_CORE(dst ,src ,ppsx ,psx ,sx ,pplx ,plx ,lx ,t0 ,t1) +DERING_CORE((%%REGa) ,(%%REGa, %1) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) +DERING_CORE((%%REGa, %1) ,(%%REGa, %1, 2),%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) +DERING_CORE((%%REGa, %1, 2),(%0, %1, 4) ,%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7) +DERING_CORE((%0, %1, 4) ,(%%REGd) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) +DERING_CORE((%%REGd) ,(%%REGd, %1) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) +DERING_CORE((%%REGd, %1) ,(%%REGd, %1, 2),%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7) +DERING_CORE((%%REGd, %1, 2),(%0, %1, 8) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7) +DERING_CORE((%0, %1, 8) ,(%%REGd, %1, 4),%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7) + + "1: \n\t" + : : "r" (src), "r" ((long)stride), "m" (c->pQPb), "m"(c->pQPb2) + : "%"REG_a, "%"REG_d, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - int y; - int min=255; - int max=0; - int avg; - uint8_t *p; - int s[10]; - const int QP2= c->QP/2 + 1; - - for(y=1; y<9; y++) - { - int x; - p= src + stride*y; - for(x=1; x<9; x++) - { - p++; - if(*p > max) max= *p; - if(*p < min) min= *p; - } - } - avg= (min + max + 1)>>1; - - if(max - min <deringThreshold) return; - - for(y=0; y<10; y++) - { - int t = 0; - - if(src[stride*y + 0] > avg) t+= 1; - if(src[stride*y + 1] > avg) t+= 2; - if(src[stride*y + 2] > avg) t+= 4; - if(src[stride*y + 3] > avg) t+= 8; - if(src[stride*y + 4] > avg) t+= 16; - if(src[stride*y + 5] > avg) t+= 32; - if(src[stride*y + 6] > avg) t+= 64; - if(src[stride*y + 7] > avg) t+= 128; - if(src[stride*y + 8] > avg) t+= 256; - if(src[stride*y + 9] > avg) t+= 512; - - t |= (~t)<<16; - t &= (t<<1) & (t>>1); - s[y] = t; - } - - for(y=1; y<9; y++) - { - int t = s[y-1] & s[y] & s[y+1]; - t|= t>>16; - s[y-1]= t; - } - - for(y=1; y<9; y++) - { - int x; - int t = s[y-1]; - - p= src + stride*y; - for(x=1; x<9; x++) - { - p++; - if(t & (1<<x)) - { - int f= (*(p-stride-1)) + 2*(*(p-stride)) + (*(p-stride+1)) - +2*(*(p -1)) + 4*(*p ) + 2*(*(p +1)) - +(*(p+stride-1)) + 2*(*(p+stride)) + (*(p+stride+1)); - f= (f + 8)>>4; + int y; + int min=255; + int max=0; + int avg; + uint8_t *p; + int s[10]; + const int QP2= c->QP/2 + 1; + + for(y=1; y<9; y++) + { + int x; + p= src + stride*y; + for(x=1; x<9; x++) + { + p++; + if(*p > max) max= *p; + if(*p < min) min= *p; + } + } + avg= (min + max + 1)>>1; + + if(max - min <deringThreshold) return; + + for(y=0; y<10; y++) + { + int t = 0; + + if(src[stride*y + 0] > avg) t+= 1; + if(src[stride*y + 1] > avg) t+= 2; + if(src[stride*y + 2] > avg) t+= 4; + if(src[stride*y + 3] > avg) t+= 8; + if(src[stride*y + 4] > avg) t+= 16; + if(src[stride*y + 5] > avg) t+= 32; + if(src[stride*y + 6] > avg) t+= 64; + if(src[stride*y + 7] > avg) t+= 128; + if(src[stride*y + 8] > avg) t+= 256; + if(src[stride*y + 9] > avg) t+= 512; + + t |= (~t)<<16; + t &= (t<<1) & (t>>1); + s[y] = t; + } + + for(y=1; y<9; y++) + { + int t = s[y-1] & s[y] & s[y+1]; + t|= t>>16; + s[y-1]= t; + } + + for(y=1; y<9; y++) + { + int x; + int t = s[y-1]; + + p= src + stride*y; + for(x=1; x<9; x++) + { + p++; + if(t & (1<<x)) + { + int f= (*(p-stride-1)) + 2*(*(p-stride)) + (*(p-stride+1)) + +2*(*(p -1)) + 4*(*p ) + 2*(*(p +1)) + +(*(p+stride-1)) + 2*(*(p+stride)) + (*(p+stride+1)); + f= (f + 8)>>4; #ifdef DEBUG_DERING_THRESHOLD - asm volatile("emms\n\t":); - { - static long long numPixels=0; - if(x!=1 && x!=8 && y!=1 && y!=8) numPixels++; -// if((max-min)<20 || (max-min)*QP<200) -// if((max-min)*QP < 500) -// if(max-min<QP/2) - if(max-min < 20) - { - static int numSkiped=0; - static int errorSum=0; - static int worstQP=0; - static int worstRange=0; - static int worstDiff=0; - int diff= (f - *p); - int absDiff= ABS(diff); - int error= diff*diff; - - if(x==1 || x==8 || y==1 || y==8) continue; - - numSkiped++; - if(absDiff > worstDiff) - { - worstDiff= absDiff; - worstQP= QP; - worstRange= max-min; - } - errorSum+= error; - - if(1024LL*1024LL*1024LL % numSkiped == 0) - { - printf( "sum:%1.3f, skip:%d, wQP:%d, " - "wRange:%d, wDiff:%d, relSkip:%1.3f\n", - (float)errorSum/numSkiped, numSkiped, worstQP, worstRange, - worstDiff, (float)numSkiped/numPixels); - } - } - } + asm volatile("emms\n\t":); + { + static long long numPixels=0; + if(x!=1 && x!=8 && y!=1 && y!=8) numPixels++; +// if((max-min)<20 || (max-min)*QP<200) +// if((max-min)*QP < 500) +// if(max-min<QP/2) + if(max-min < 20) + { + static int numSkiped=0; + static int errorSum=0; + static int worstQP=0; + static int worstRange=0; + static int worstDiff=0; + int diff= (f - *p); + int absDiff= ABS(diff); + int error= diff*diff; + + if(x==1 || x==8 || y==1 || y==8) continue; + + numSkiped++; + if(absDiff > worstDiff) + { + worstDiff= absDiff; + worstQP= QP; + worstRange= max-min; + } + errorSum+= error; + + if(1024LL*1024LL*1024LL % numSkiped == 0) + { + printf( "sum:%1.3f, skip:%d, wQP:%d, " + "wRange:%d, wDiff:%d, relSkip:%1.3f\n", + (float)errorSum/numSkiped, numSkiped, worstQP, worstRange, + worstDiff, (float)numSkiped/numPixels); + } + } + } #endif - if (*p + QP2 < f) *p= *p + QP2; - else if(*p - QP2 > f) *p= *p - QP2; - else *p=f; - } - } - } + if (*p + QP2 < f) *p= *p + QP2; + else if(*p - QP2 > f) *p= *p - QP2; + else *p=f; + } + } + } #ifdef DEBUG_DERING_THRESHOLD - if(max-min < 20) - { - for(y=1; y<9; y++) - { - int x; - int t = 0; - p= src + stride*y; - for(x=1; x<9; x++) - { - p++; - *p = MIN(*p + 20, 255); - } - } -// src[0] = src[7]=src[stride*7]=src[stride*7 + 7]=255; - } + if(max-min < 20) + { + for(y=1; y<9; y++) + { + int x; + int t = 0; + p= src + stride*y; + for(x=1; x<9; x++) + { + p++; + *p = MIN(*p + 20, 255); + } + } +// src[0] = src[7]=src[stride*7]=src[stride*7 + 7]=255; + } #endif #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -1549,46 +1549,46 @@ static inline void RENAME(deInterlaceInterpolateLinear)(uint8_t src[], int stride) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= 4*stride; - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 - - "movq (%0), %%mm0 \n\t" - "movq (%%"REG_a", %1), %%mm1 \n\t" - PAVGB(%%mm1, %%mm0) - "movq %%mm0, (%%"REG_a") \n\t" - "movq (%0, %1, 4), %%mm0 \n\t" - PAVGB(%%mm0, %%mm1) - "movq %%mm1, (%%"REG_a", %1, 2) \n\t" - "movq (%%"REG_c", %1), %%mm1 \n\t" - PAVGB(%%mm1, %%mm0) - "movq %%mm0, (%%"REG_c") \n\t" - "movq (%0, %1, 8), %%mm0 \n\t" - PAVGB(%%mm0, %%mm1) - "movq %%mm1, (%%"REG_c", %1, 2) \n\t" - - : : "r" (src), "r" ((long)stride) - : "%"REG_a, "%"REG_c - ); + src+= 4*stride; + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_c" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %0+8%1 ecx+4%1 + + "movq (%0), %%mm0 \n\t" + "movq (%%"REG_a", %1), %%mm1 \n\t" + PAVGB(%%mm1, %%mm0) + "movq %%mm0, (%%"REG_a") \n\t" + "movq (%0, %1, 4), %%mm0 \n\t" + PAVGB(%%mm0, %%mm1) + "movq %%mm1, (%%"REG_a", %1, 2) \n\t" + "movq (%%"REG_c", %1), %%mm1 \n\t" + PAVGB(%%mm1, %%mm0) + "movq %%mm0, (%%"REG_c") \n\t" + "movq (%0, %1, 8), %%mm0 \n\t" + PAVGB(%%mm0, %%mm1) + "movq %%mm1, (%%"REG_c", %1, 2) \n\t" + + : : "r" (src), "r" ((long)stride) + : "%"REG_a, "%"REG_c + ); #else - int a, b, x; - src+= 4*stride; - - for(x=0; x<2; x++){ - a= *(uint32_t*)&src[stride*0]; - b= *(uint32_t*)&src[stride*2]; - *(uint32_t*)&src[stride*1]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - a= *(uint32_t*)&src[stride*4]; - *(uint32_t*)&src[stride*3]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - b= *(uint32_t*)&src[stride*6]; - *(uint32_t*)&src[stride*5]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - a= *(uint32_t*)&src[stride*8]; - *(uint32_t*)&src[stride*7]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - src += 4; - } + int a, b, x; + src+= 4*stride; + + for(x=0; x<2; x++){ + a= *(uint32_t*)&src[stride*0]; + b= *(uint32_t*)&src[stride*2]; + *(uint32_t*)&src[stride*1]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + a= *(uint32_t*)&src[stride*4]; + *(uint32_t*)&src[stride*3]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + b= *(uint32_t*)&src[stride*6]; + *(uint32_t*)&src[stride*5]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + a= *(uint32_t*)&src[stride*8]; + *(uint32_t*)&src[stride*7]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + src += 4; + } #endif } @@ -1602,58 +1602,58 @@ static inline void RENAME(deInterlaceInterpolateCubic)(uint8_t src[], int stride) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*3; - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" - "lea (%%"REG_d", %1, 4), %%"REG_c" \n\t" - "add %1, %%"REG_c" \n\t" - "pxor %%mm7, %%mm7 \n\t" -// 0 1 2 3 4 5 6 7 8 9 10 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx + src+= stride*3; + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" + "lea (%%"REG_d", %1, 4), %%"REG_c" \n\t" + "add %1, %%"REG_c" \n\t" + "pxor %%mm7, %%mm7 \n\t" +// 0 1 2 3 4 5 6 7 8 9 10 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx #define REAL_DEINT_CUBIC(a,b,c,d,e)\ - "movq " #a ", %%mm0 \n\t"\ - "movq " #b ", %%mm1 \n\t"\ - "movq " #d ", %%mm2 \n\t"\ - "movq " #e ", %%mm3 \n\t"\ - PAVGB(%%mm2, %%mm1) /* (b+d) /2 */\ - PAVGB(%%mm3, %%mm0) /* a(a+e) /2 */\ - "movq %%mm0, %%mm2 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm2 \n\t"\ - "movq %%mm1, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "psubw %%mm1, %%mm0 \n\t" /* L(a+e - (b+d))/2 */\ - "psubw %%mm3, %%mm2 \n\t" /* H(a+e - (b+d))/2 */\ - "psraw $3, %%mm0 \n\t" /* L(a+e - (b+d))/16 */\ - "psraw $3, %%mm2 \n\t" /* H(a+e - (b+d))/16 */\ - "psubw %%mm0, %%mm1 \n\t" /* L(9b + 9d - a - e)/16 */\ - "psubw %%mm2, %%mm3 \n\t" /* H(9b + 9d - a - e)/16 */\ - "packuswb %%mm3, %%mm1 \n\t"\ - "movq %%mm1, " #c " \n\t" + "movq " #a ", %%mm0 \n\t"\ + "movq " #b ", %%mm1 \n\t"\ + "movq " #d ", %%mm2 \n\t"\ + "movq " #e ", %%mm3 \n\t"\ + PAVGB(%%mm2, %%mm1) /* (b+d) /2 */\ + PAVGB(%%mm3, %%mm0) /* a(a+e) /2 */\ + "movq %%mm0, %%mm2 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm2 \n\t"\ + "movq %%mm1, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "psubw %%mm1, %%mm0 \n\t" /* L(a+e - (b+d))/2 */\ + "psubw %%mm3, %%mm2 \n\t" /* H(a+e - (b+d))/2 */\ + "psraw $3, %%mm0 \n\t" /* L(a+e - (b+d))/16 */\ + "psraw $3, %%mm2 \n\t" /* H(a+e - (b+d))/16 */\ + "psubw %%mm0, %%mm1 \n\t" /* L(9b + 9d - a - e)/16 */\ + "psubw %%mm2, %%mm3 \n\t" /* H(9b + 9d - a - e)/16 */\ + "packuswb %%mm3, %%mm1 \n\t"\ + "movq %%mm1, " #c " \n\t" #define DEINT_CUBIC(a,b,c,d,e) REAL_DEINT_CUBIC(a,b,c,d,e) -DEINT_CUBIC((%0), (%%REGa, %1), (%%REGa, %1, 2), (%0, %1, 4), (%%REGd, %1)) -DEINT_CUBIC((%%REGa, %1), (%0, %1, 4), (%%REGd), (%%REGd, %1), (%0, %1, 8)) -DEINT_CUBIC((%0, %1, 4), (%%REGd, %1), (%%REGd, %1, 2), (%0, %1, 8), (%%REGc)) -DEINT_CUBIC((%%REGd, %1), (%0, %1, 8), (%%REGd, %1, 4), (%%REGc), (%%REGc, %1, 2)) - - : : "r" (src), "r" ((long)stride) - : "%"REG_a, "%"REG_d, "%"REG_c - ); +DEINT_CUBIC((%0) , (%%REGa, %1), (%%REGa, %1, 2), (%0, %1, 4) , (%%REGd, %1)) +DEINT_CUBIC((%%REGa, %1), (%0, %1, 4) , (%%REGd) , (%%REGd, %1), (%0, %1, 8)) +DEINT_CUBIC((%0, %1, 4) , (%%REGd, %1), (%%REGd, %1, 2), (%0, %1, 8) , (%%REGc)) +DEINT_CUBIC((%%REGd, %1), (%0, %1, 8) , (%%REGd, %1, 4), (%%REGc) , (%%REGc, %1, 2)) + + : : "r" (src), "r" ((long)stride) + : "%"REG_a, "%"REG_d, "%"REG_c + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - int x; - src+= stride*3; - for(x=0; x<8; x++) - { - src[stride*3] = CLIP((-src[0] + 9*src[stride*2] + 9*src[stride*4] - src[stride*6])>>4); - src[stride*5] = CLIP((-src[stride*2] + 9*src[stride*4] + 9*src[stride*6] - src[stride*8])>>4); - src[stride*7] = CLIP((-src[stride*4] + 9*src[stride*6] + 9*src[stride*8] - src[stride*10])>>4); - src[stride*9] = CLIP((-src[stride*6] + 9*src[stride*8] + 9*src[stride*10] - src[stride*12])>>4); - src++; - } + int x; + src+= stride*3; + for(x=0; x<8; x++) + { + src[stride*3] = CLIP((-src[0] + 9*src[stride*2] + 9*src[stride*4] - src[stride*6])>>4); + src[stride*5] = CLIP((-src[stride*2] + 9*src[stride*4] + 9*src[stride*6] - src[stride*8])>>4); + src[stride*7] = CLIP((-src[stride*4] + 9*src[stride*6] + 9*src[stride*8] - src[stride*10])>>4); + src[stride*9] = CLIP((-src[stride*6] + 9*src[stride*8] + 9*src[stride*10] - src[stride*12])>>4); + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -1667,73 +1667,73 @@ static inline void RENAME(deInterlaceFF)(uint8_t src[], int stride, uint8_t *tmp) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*4; - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" - "pxor %%mm7, %%mm7 \n\t" - "movq (%2), %%mm0 \n\t" -// 0 1 2 3 4 5 6 7 8 9 10 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx + src+= stride*4; + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" + "pxor %%mm7, %%mm7 \n\t" + "movq (%2), %%mm0 \n\t" +// 0 1 2 3 4 5 6 7 8 9 10 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx #define REAL_DEINT_FF(a,b,c,d)\ - "movq " #a ", %%mm1 \n\t"\ - "movq " #b ", %%mm2 \n\t"\ - "movq " #c ", %%mm3 \n\t"\ - "movq " #d ", %%mm4 \n\t"\ - PAVGB(%%mm3, %%mm1) \ - PAVGB(%%mm4, %%mm0) \ - "movq %%mm0, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm0 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "movq %%mm1, %%mm4 \n\t"\ - "punpcklbw %%mm7, %%mm1 \n\t"\ - "punpckhbw %%mm7, %%mm4 \n\t"\ - "psllw $2, %%mm1 \n\t"\ - "psllw $2, %%mm4 \n\t"\ - "psubw %%mm0, %%mm1 \n\t"\ - "psubw %%mm3, %%mm4 \n\t"\ - "movq %%mm2, %%mm5 \n\t"\ - "movq %%mm2, %%mm0 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm5 \n\t"\ - "paddw %%mm2, %%mm1 \n\t"\ - "paddw %%mm5, %%mm4 \n\t"\ - "psraw $2, %%mm1 \n\t"\ - "psraw $2, %%mm4 \n\t"\ - "packuswb %%mm4, %%mm1 \n\t"\ - "movq %%mm1, " #b " \n\t"\ + "movq " #a ", %%mm1 \n\t"\ + "movq " #b ", %%mm2 \n\t"\ + "movq " #c ", %%mm3 \n\t"\ + "movq " #d ", %%mm4 \n\t"\ + PAVGB(%%mm3, %%mm1) \ + PAVGB(%%mm4, %%mm0) \ + "movq %%mm0, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm0 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "movq %%mm1, %%mm4 \n\t"\ + "punpcklbw %%mm7, %%mm1 \n\t"\ + "punpckhbw %%mm7, %%mm4 \n\t"\ + "psllw $2, %%mm1 \n\t"\ + "psllw $2, %%mm4 \n\t"\ + "psubw %%mm0, %%mm1 \n\t"\ + "psubw %%mm3, %%mm4 \n\t"\ + "movq %%mm2, %%mm5 \n\t"\ + "movq %%mm2, %%mm0 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm5 \n\t"\ + "paddw %%mm2, %%mm1 \n\t"\ + "paddw %%mm5, %%mm4 \n\t"\ + "psraw $2, %%mm1 \n\t"\ + "psraw $2, %%mm4 \n\t"\ + "packuswb %%mm4, %%mm1 \n\t"\ + "movq %%mm1, " #b " \n\t"\ #define DEINT_FF(a,b,c,d) REAL_DEINT_FF(a,b,c,d) -DEINT_FF((%0) , (%%REGa) , (%%REGa, %1), (%%REGa, %1, 2)) -DEINT_FF((%%REGa, %1), (%%REGa, %1, 2), (%0, %1, 4), (%%REGd) ) -DEINT_FF((%0, %1, 4), (%%REGd) , (%%REGd, %1), (%%REGd, %1, 2)) -DEINT_FF((%%REGd, %1), (%%REGd, %1, 2), (%0, %1, 8), (%%REGd, %1, 4)) - - "movq %%mm0, (%2) \n\t" - : : "r" (src), "r" ((long)stride), "r"(tmp) - : "%"REG_a, "%"REG_d - ); +DEINT_FF((%0) , (%%REGa) , (%%REGa, %1), (%%REGa, %1, 2)) +DEINT_FF((%%REGa, %1), (%%REGa, %1, 2), (%0, %1, 4) , (%%REGd) ) +DEINT_FF((%0, %1, 4) , (%%REGd) , (%%REGd, %1), (%%REGd, %1, 2)) +DEINT_FF((%%REGd, %1), (%%REGd, %1, 2), (%0, %1, 8) , (%%REGd, %1, 4)) + + "movq %%mm0, (%2) \n\t" + : : "r" (src), "r" ((long)stride), "r"(tmp) + : "%"REG_a, "%"REG_d + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - int x; - src+= stride*4; - for(x=0; x<8; x++) - { - int t1= tmp[x]; - int t2= src[stride*1]; - - src[stride*1]= CLIP((-t1 + 4*src[stride*0] + 2*t2 + 4*src[stride*2] - src[stride*3] + 4)>>3); - t1= src[stride*4]; - src[stride*3]= CLIP((-t2 + 4*src[stride*2] + 2*t1 + 4*src[stride*4] - src[stride*5] + 4)>>3); - t2= src[stride*6]; - src[stride*5]= CLIP((-t1 + 4*src[stride*4] + 2*t2 + 4*src[stride*6] - src[stride*7] + 4)>>3); - t1= src[stride*8]; - src[stride*7]= CLIP((-t2 + 4*src[stride*6] + 2*t1 + 4*src[stride*8] - src[stride*9] + 4)>>3); - tmp[x]= t1; - - src++; - } + int x; + src+= stride*4; + for(x=0; x<8; x++) + { + int t1= tmp[x]; + int t2= src[stride*1]; + + src[stride*1]= CLIP((-t1 + 4*src[stride*0] + 2*t2 + 4*src[stride*2] - src[stride*3] + 4)>>3); + t1= src[stride*4]; + src[stride*3]= CLIP((-t2 + 4*src[stride*2] + 2*t1 + 4*src[stride*4] - src[stride*5] + 4)>>3); + t2= src[stride*6]; + src[stride*5]= CLIP((-t1 + 4*src[stride*4] + 2*t2 + 4*src[stride*6] - src[stride*7] + 4)>>3); + t1= src[stride*8]; + src[stride*7]= CLIP((-t2 + 4*src[stride*6] + 2*t1 + 4*src[stride*8] - src[stride*9] + 4)>>3); + tmp[x]= t1; + + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -1747,48 +1747,48 @@ static inline void RENAME(deInterlaceL5)(uint8_t src[], int stride, uint8_t *tmp, uint8_t *tmp2) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= stride*4; - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" - "pxor %%mm7, %%mm7 \n\t" - "movq (%2), %%mm0 \n\t" - "movq (%3), %%mm1 \n\t" -// 0 1 2 3 4 5 6 7 8 9 10 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx + src+= stride*4; + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" + "pxor %%mm7, %%mm7 \n\t" + "movq (%2), %%mm0 \n\t" + "movq (%3), %%mm1 \n\t" +// 0 1 2 3 4 5 6 7 8 9 10 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 ecx #define REAL_DEINT_L5(t1,t2,a,b,c)\ - "movq " #a ", %%mm2 \n\t"\ - "movq " #b ", %%mm3 \n\t"\ - "movq " #c ", %%mm4 \n\t"\ - PAVGB(t2, %%mm3) \ - PAVGB(t1, %%mm4) \ - "movq %%mm2, %%mm5 \n\t"\ - "movq %%mm2, " #t1 " \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm5 \n\t"\ - "movq %%mm2, %%mm6 \n\t"\ - "paddw %%mm2, %%mm2 \n\t"\ - "paddw %%mm6, %%mm2 \n\t"\ - "movq %%mm5, %%mm6 \n\t"\ - "paddw %%mm5, %%mm5 \n\t"\ - "paddw %%mm6, %%mm5 \n\t"\ - "movq %%mm3, %%mm6 \n\t"\ - "punpcklbw %%mm7, %%mm3 \n\t"\ - "punpckhbw %%mm7, %%mm6 \n\t"\ - "paddw %%mm3, %%mm3 \n\t"\ - "paddw %%mm6, %%mm6 \n\t"\ - "paddw %%mm3, %%mm2 \n\t"\ - "paddw %%mm6, %%mm5 \n\t"\ - "movq %%mm4, %%mm6 \n\t"\ - "punpcklbw %%mm7, %%mm4 \n\t"\ - "punpckhbw %%mm7, %%mm6 \n\t"\ - "psubw %%mm4, %%mm2 \n\t"\ - "psubw %%mm6, %%mm5 \n\t"\ - "psraw $2, %%mm2 \n\t"\ - "psraw $2, %%mm5 \n\t"\ - "packuswb %%mm5, %%mm2 \n\t"\ - "movq %%mm2, " #a " \n\t"\ + "movq " #a ", %%mm2 \n\t"\ + "movq " #b ", %%mm3 \n\t"\ + "movq " #c ", %%mm4 \n\t"\ + PAVGB(t2, %%mm3) \ + PAVGB(t1, %%mm4) \ + "movq %%mm2, %%mm5 \n\t"\ + "movq %%mm2, " #t1 " \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm5 \n\t"\ + "movq %%mm2, %%mm6 \n\t"\ + "paddw %%mm2, %%mm2 \n\t"\ + "paddw %%mm6, %%mm2 \n\t"\ + "movq %%mm5, %%mm6 \n\t"\ + "paddw %%mm5, %%mm5 \n\t"\ + "paddw %%mm6, %%mm5 \n\t"\ + "movq %%mm3, %%mm6 \n\t"\ + "punpcklbw %%mm7, %%mm3 \n\t"\ + "punpckhbw %%mm7, %%mm6 \n\t"\ + "paddw %%mm3, %%mm3 \n\t"\ + "paddw %%mm6, %%mm6 \n\t"\ + "paddw %%mm3, %%mm2 \n\t"\ + "paddw %%mm6, %%mm5 \n\t"\ + "movq %%mm4, %%mm6 \n\t"\ + "punpcklbw %%mm7, %%mm4 \n\t"\ + "punpckhbw %%mm7, %%mm6 \n\t"\ + "psubw %%mm4, %%mm2 \n\t"\ + "psubw %%mm6, %%mm5 \n\t"\ + "psraw $2, %%mm2 \n\t"\ + "psraw $2, %%mm5 \n\t"\ + "packuswb %%mm5, %%mm2 \n\t"\ + "movq %%mm2, " #a " \n\t"\ #define DEINT_L5(t1,t2,a,b,c) REAL_DEINT_L5(t1,t2,a,b,c) @@ -1801,41 +1801,41 @@ DEINT_L5(%%mm0, %%mm1, (%%REGd, %1) , (%%REGd, %1, 2), (%0, %1, 8) ) DEINT_L5(%%mm1, %%mm0, (%%REGd, %1, 2), (%0, %1, 8) , (%%REGd, %1, 4)) - "movq %%mm0, (%2) \n\t" - "movq %%mm1, (%3) \n\t" - : : "r" (src), "r" ((long)stride), "r"(tmp), "r"(tmp2) - : "%"REG_a, "%"REG_d - ); + "movq %%mm0, (%2) \n\t" + "movq %%mm1, (%3) \n\t" + : : "r" (src), "r" ((long)stride), "r"(tmp), "r"(tmp2) + : "%"REG_a, "%"REG_d + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - int x; - src+= stride*4; - for(x=0; x<8; x++) - { - int t1= tmp[x]; - int t2= tmp2[x]; - int t3= src[0]; - - src[stride*0]= CLIP((-(t1 + src[stride*2]) + 2*(t2 + src[stride*1]) + 6*t3 + 4)>>3); - t1= src[stride*1]; - src[stride*1]= CLIP((-(t2 + src[stride*3]) + 2*(t3 + src[stride*2]) + 6*t1 + 4)>>3); - t2= src[stride*2]; - src[stride*2]= CLIP((-(t3 + src[stride*4]) + 2*(t1 + src[stride*3]) + 6*t2 + 4)>>3); - t3= src[stride*3]; - src[stride*3]= CLIP((-(t1 + src[stride*5]) + 2*(t2 + src[stride*4]) + 6*t3 + 4)>>3); - t1= src[stride*4]; - src[stride*4]= CLIP((-(t2 + src[stride*6]) + 2*(t3 + src[stride*5]) + 6*t1 + 4)>>3); - t2= src[stride*5]; - src[stride*5]= CLIP((-(t3 + src[stride*7]) + 2*(t1 + src[stride*6]) + 6*t2 + 4)>>3); - t3= src[stride*6]; - src[stride*6]= CLIP((-(t1 + src[stride*8]) + 2*(t2 + src[stride*7]) + 6*t3 + 4)>>3); - t1= src[stride*7]; - src[stride*7]= CLIP((-(t2 + src[stride*9]) + 2*(t3 + src[stride*8]) + 6*t1 + 4)>>3); - - tmp[x]= t3; - tmp2[x]= t1; - - src++; - } + int x; + src+= stride*4; + for(x=0; x<8; x++) + { + int t1= tmp[x]; + int t2= tmp2[x]; + int t3= src[0]; + + src[stride*0]= CLIP((-(t1 + src[stride*2]) + 2*(t2 + src[stride*1]) + 6*t3 + 4)>>3); + t1= src[stride*1]; + src[stride*1]= CLIP((-(t2 + src[stride*3]) + 2*(t3 + src[stride*2]) + 6*t1 + 4)>>3); + t2= src[stride*2]; + src[stride*2]= CLIP((-(t3 + src[stride*4]) + 2*(t1 + src[stride*3]) + 6*t2 + 4)>>3); + t3= src[stride*3]; + src[stride*3]= CLIP((-(t1 + src[stride*5]) + 2*(t2 + src[stride*4]) + 6*t3 + 4)>>3); + t1= src[stride*4]; + src[stride*4]= CLIP((-(t2 + src[stride*6]) + 2*(t3 + src[stride*5]) + 6*t1 + 4)>>3); + t2= src[stride*5]; + src[stride*5]= CLIP((-(t3 + src[stride*7]) + 2*(t1 + src[stride*6]) + 6*t2 + 4)>>3); + t3= src[stride*6]; + src[stride*6]= CLIP((-(t1 + src[stride*8]) + 2*(t2 + src[stride*7]) + 6*t3 + 4)>>3); + t1= src[stride*7]; + src[stride*7]= CLIP((-(t2 + src[stride*9]) + 2*(t3 + src[stride*8]) + 6*t1 + 4)>>3); + + tmp[x]= t3; + tmp2[x]= t1; + + src++; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -1849,95 +1849,95 @@ static inline void RENAME(deInterlaceBlendLinear)(uint8_t src[], int stride, uint8_t *tmp) { #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - src+= 4*stride; - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 - - "movq (%2), %%mm0 \n\t" // L0 - "movq (%%"REG_a"), %%mm1 \n\t" // L2 - PAVGB(%%mm1, %%mm0) // L0+L2 - "movq (%0), %%mm2 \n\t" // L1 - PAVGB(%%mm2, %%mm0) - "movq %%mm0, (%0) \n\t" - "movq (%%"REG_a", %1), %%mm0 \n\t" // L3 - PAVGB(%%mm0, %%mm2) // L1+L3 - PAVGB(%%mm1, %%mm2) // 2L2 + L1 + L3 - "movq %%mm2, (%%"REG_a") \n\t" - "movq (%%"REG_a", %1, 2), %%mm2 \n\t" // L4 - PAVGB(%%mm2, %%mm1) // L2+L4 - PAVGB(%%mm0, %%mm1) // 2L3 + L2 + L4 - "movq %%mm1, (%%"REG_a", %1) \n\t" - "movq (%0, %1, 4), %%mm1 \n\t" // L5 - PAVGB(%%mm1, %%mm0) // L3+L5 - PAVGB(%%mm2, %%mm0) // 2L4 + L3 + L5 - "movq %%mm0, (%%"REG_a", %1, 2) \n\t" - "movq (%%"REG_d"), %%mm0 \n\t" // L6 - PAVGB(%%mm0, %%mm2) // L4+L6 - PAVGB(%%mm1, %%mm2) // 2L5 + L4 + L6 - "movq %%mm2, (%0, %1, 4) \n\t" - "movq (%%"REG_d", %1), %%mm2 \n\t" // L7 - PAVGB(%%mm2, %%mm1) // L5+L7 - PAVGB(%%mm0, %%mm1) // 2L6 + L5 + L7 - "movq %%mm1, (%%"REG_d") \n\t" - "movq (%%"REG_d", %1, 2), %%mm1 \n\t" // L8 - PAVGB(%%mm1, %%mm0) // L6+L8 - PAVGB(%%mm2, %%mm0) // 2L7 + L6 + L8 - "movq %%mm0, (%%"REG_d", %1) \n\t" - "movq (%0, %1, 8), %%mm0 \n\t" // L9 - PAVGB(%%mm0, %%mm2) // L7+L9 - PAVGB(%%mm1, %%mm2) // 2L8 + L7 + L9 - "movq %%mm2, (%%"REG_d", %1, 2) \n\t" - "movq %%mm1, (%2) \n\t" - - : : "r" (src), "r" ((long)stride), "r" (tmp) - : "%"REG_a, "%"REG_d - ); + src+= 4*stride; + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + + "movq (%2), %%mm0 \n\t" // L0 + "movq (%%"REG_a"), %%mm1 \n\t" // L2 + PAVGB(%%mm1, %%mm0) // L0+L2 + "movq (%0), %%mm2 \n\t" // L1 + PAVGB(%%mm2, %%mm0) + "movq %%mm0, (%0) \n\t" + "movq (%%"REG_a", %1), %%mm0 \n\t" // L3 + PAVGB(%%mm0, %%mm2) // L1+L3 + PAVGB(%%mm1, %%mm2) // 2L2 + L1 + L3 + "movq %%mm2, (%%"REG_a") \n\t" + "movq (%%"REG_a", %1, 2), %%mm2 \n\t" // L4 + PAVGB(%%mm2, %%mm1) // L2+L4 + PAVGB(%%mm0, %%mm1) // 2L3 + L2 + L4 + "movq %%mm1, (%%"REG_a", %1) \n\t" + "movq (%0, %1, 4), %%mm1 \n\t" // L5 + PAVGB(%%mm1, %%mm0) // L3+L5 + PAVGB(%%mm2, %%mm0) // 2L4 + L3 + L5 + "movq %%mm0, (%%"REG_a", %1, 2) \n\t" + "movq (%%"REG_d"), %%mm0 \n\t" // L6 + PAVGB(%%mm0, %%mm2) // L4+L6 + PAVGB(%%mm1, %%mm2) // 2L5 + L4 + L6 + "movq %%mm2, (%0, %1, 4) \n\t" + "movq (%%"REG_d", %1), %%mm2 \n\t" // L7 + PAVGB(%%mm2, %%mm1) // L5+L7 + PAVGB(%%mm0, %%mm1) // 2L6 + L5 + L7 + "movq %%mm1, (%%"REG_d") \n\t" + "movq (%%"REG_d", %1, 2), %%mm1 \n\t" // L8 + PAVGB(%%mm1, %%mm0) // L6+L8 + PAVGB(%%mm2, %%mm0) // 2L7 + L6 + L8 + "movq %%mm0, (%%"REG_d", %1) \n\t" + "movq (%0, %1, 8), %%mm0 \n\t" // L9 + PAVGB(%%mm0, %%mm2) // L7+L9 + PAVGB(%%mm1, %%mm2) // 2L8 + L7 + L9 + "movq %%mm2, (%%"REG_d", %1, 2) \n\t" + "movq %%mm1, (%2) \n\t" + + : : "r" (src), "r" ((long)stride), "r" (tmp) + : "%"REG_a, "%"REG_d + ); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) - int a, b, c, x; - src+= 4*stride; - - for(x=0; x<2; x++){ - a= *(uint32_t*)&tmp[stride*0]; - b= *(uint32_t*)&src[stride*0]; - c= *(uint32_t*)&src[stride*1]; - a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*0]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - - a= *(uint32_t*)&src[stride*2]; - b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*1]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); - - b= *(uint32_t*)&src[stride*3]; - c= (b&c) + (((b^c)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*2]= (c|a) - (((c^a)&0xFEFEFEFEUL)>>1); - - c= *(uint32_t*)&src[stride*4]; - a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*3]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - - a= *(uint32_t*)&src[stride*5]; - b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*4]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); - - b= *(uint32_t*)&src[stride*6]; - c= (b&c) + (((b^c)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*5]= (c|a) - (((c^a)&0xFEFEFEFEUL)>>1); - - c= *(uint32_t*)&src[stride*7]; - a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*6]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); - - a= *(uint32_t*)&src[stride*8]; - b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); - *(uint32_t*)&src[stride*7]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); - - *(uint32_t*)&tmp[stride*0]= c; - src += 4; - tmp += 4; - } + int a, b, c, x; + src+= 4*stride; + + for(x=0; x<2; x++){ + a= *(uint32_t*)&tmp[stride*0]; + b= *(uint32_t*)&src[stride*0]; + c= *(uint32_t*)&src[stride*1]; + a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*0]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + + a= *(uint32_t*)&src[stride*2]; + b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*1]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); + + b= *(uint32_t*)&src[stride*3]; + c= (b&c) + (((b^c)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*2]= (c|a) - (((c^a)&0xFEFEFEFEUL)>>1); + + c= *(uint32_t*)&src[stride*4]; + a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*3]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + + a= *(uint32_t*)&src[stride*5]; + b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*4]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); + + b= *(uint32_t*)&src[stride*6]; + c= (b&c) + (((b^c)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*5]= (c|a) - (((c^a)&0xFEFEFEFEUL)>>1); + + c= *(uint32_t*)&src[stride*7]; + a= (a&c) + (((a^c)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*6]= (a|b) - (((a^b)&0xFEFEFEFEUL)>>1); + + a= *(uint32_t*)&src[stride*8]; + b= (a&b) + (((a^b)&0xFEFEFEFEUL)>>1); + *(uint32_t*)&src[stride*7]= (c|b) - (((c^b)&0xFEFEFEFEUL)>>1); + + *(uint32_t*)&tmp[stride*0]= c; + src += 4; + tmp += 4; + } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -1950,119 +1950,119 @@ static inline void RENAME(deInterlaceMedian)(uint8_t src[], int stride) { #ifdef HAVE_MMX - src+= 4*stride; + src+= 4*stride; #ifdef HAVE_MMX2 - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 - - "movq (%0), %%mm0 \n\t" // - "movq (%%"REG_a", %1), %%mm2 \n\t" // - "movq (%%"REG_a"), %%mm1 \n\t" // - "movq %%mm0, %%mm3 \n\t" - "pmaxub %%mm1, %%mm0 \n\t" // - "pminub %%mm3, %%mm1 \n\t" // - "pmaxub %%mm2, %%mm1 \n\t" // - "pminub %%mm1, %%mm0 \n\t" - "movq %%mm0, (%%"REG_a") \n\t" - - "movq (%0, %1, 4), %%mm0 \n\t" // - "movq (%%"REG_a", %1, 2), %%mm1 \n\t" // - "movq %%mm2, %%mm3 \n\t" - "pmaxub %%mm1, %%mm2 \n\t" // - "pminub %%mm3, %%mm1 \n\t" // - "pmaxub %%mm0, %%mm1 \n\t" // - "pminub %%mm1, %%mm2 \n\t" - "movq %%mm2, (%%"REG_a", %1, 2) \n\t" - - "movq (%%"REG_d"), %%mm2 \n\t" // - "movq (%%"REG_d", %1), %%mm1 \n\t" // - "movq %%mm2, %%mm3 \n\t" - "pmaxub %%mm0, %%mm2 \n\t" // - "pminub %%mm3, %%mm0 \n\t" // - "pmaxub %%mm1, %%mm0 \n\t" // - "pminub %%mm0, %%mm2 \n\t" - "movq %%mm2, (%%"REG_d") \n\t" - - "movq (%%"REG_d", %1, 2), %%mm2 \n\t" // - "movq (%0, %1, 8), %%mm0 \n\t" // - "movq %%mm2, %%mm3 \n\t" - "pmaxub %%mm0, %%mm2 \n\t" // - "pminub %%mm3, %%mm0 \n\t" // - "pmaxub %%mm1, %%mm0 \n\t" // - "pminub %%mm0, %%mm2 \n\t" - "movq %%mm2, (%%"REG_d", %1, 2) \n\t" - - - : : "r" (src), "r" ((long)stride) - : "%"REG_a, "%"REG_d - ); + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + + "movq (%0), %%mm0 \n\t" // + "movq (%%"REG_a", %1), %%mm2 \n\t" // + "movq (%%"REG_a"), %%mm1 \n\t" // + "movq %%mm0, %%mm3 \n\t" + "pmaxub %%mm1, %%mm0 \n\t" // + "pminub %%mm3, %%mm1 \n\t" // + "pmaxub %%mm2, %%mm1 \n\t" // + "pminub %%mm1, %%mm0 \n\t" + "movq %%mm0, (%%"REG_a") \n\t" + + "movq (%0, %1, 4), %%mm0 \n\t" // + "movq (%%"REG_a", %1, 2), %%mm1 \n\t" // + "movq %%mm2, %%mm3 \n\t" + "pmaxub %%mm1, %%mm2 \n\t" // + "pminub %%mm3, %%mm1 \n\t" // + "pmaxub %%mm0, %%mm1 \n\t" // + "pminub %%mm1, %%mm2 \n\t" + "movq %%mm2, (%%"REG_a", %1, 2) \n\t" + + "movq (%%"REG_d"), %%mm2 \n\t" // + "movq (%%"REG_d", %1), %%mm1 \n\t" // + "movq %%mm2, %%mm3 \n\t" + "pmaxub %%mm0, %%mm2 \n\t" // + "pminub %%mm3, %%mm0 \n\t" // + "pmaxub %%mm1, %%mm0 \n\t" // + "pminub %%mm0, %%mm2 \n\t" + "movq %%mm2, (%%"REG_d") \n\t" + + "movq (%%"REG_d", %1, 2), %%mm2 \n\t" // + "movq (%0, %1, 8), %%mm0 \n\t" // + "movq %%mm2, %%mm3 \n\t" + "pmaxub %%mm0, %%mm2 \n\t" // + "pminub %%mm3, %%mm0 \n\t" // + "pmaxub %%mm1, %%mm0 \n\t" // + "pminub %%mm0, %%mm2 \n\t" + "movq %%mm2, (%%"REG_d", %1, 2) \n\t" + + + : : "r" (src), "r" ((long)stride) + : "%"REG_a, "%"REG_d + ); #else // MMX without MMX2 - asm volatile( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 - "pxor %%mm7, %%mm7 \n\t" + asm volatile( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a", %1, 4), %%"REG_d" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + "pxor %%mm7, %%mm7 \n\t" #define REAL_MEDIAN(a,b,c)\ - "movq " #a ", %%mm0 \n\t"\ - "movq " #b ", %%mm2 \n\t"\ - "movq " #c ", %%mm1 \n\t"\ - "movq %%mm0, %%mm3 \n\t"\ - "movq %%mm1, %%mm4 \n\t"\ - "movq %%mm2, %%mm5 \n\t"\ - "psubusb %%mm1, %%mm3 \n\t"\ - "psubusb %%mm2, %%mm4 \n\t"\ - "psubusb %%mm0, %%mm5 \n\t"\ - "pcmpeqb %%mm7, %%mm3 \n\t"\ - "pcmpeqb %%mm7, %%mm4 \n\t"\ - "pcmpeqb %%mm7, %%mm5 \n\t"\ - "movq %%mm3, %%mm6 \n\t"\ - "pxor %%mm4, %%mm3 \n\t"\ - "pxor %%mm5, %%mm4 \n\t"\ - "pxor %%mm6, %%mm5 \n\t"\ - "por %%mm3, %%mm1 \n\t"\ - "por %%mm4, %%mm2 \n\t"\ - "por %%mm5, %%mm0 \n\t"\ - "pand %%mm2, %%mm0 \n\t"\ - "pand %%mm1, %%mm0 \n\t"\ - "movq %%mm0, " #b " \n\t" + "movq " #a ", %%mm0 \n\t"\ + "movq " #b ", %%mm2 \n\t"\ + "movq " #c ", %%mm1 \n\t"\ + "movq %%mm0, %%mm3 \n\t"\ + "movq %%mm1, %%mm4 \n\t"\ + "movq %%mm2, %%mm5 \n\t"\ + "psubusb %%mm1, %%mm3 \n\t"\ + "psubusb %%mm2, %%mm4 \n\t"\ + "psubusb %%mm0, %%mm5 \n\t"\ + "pcmpeqb %%mm7, %%mm3 \n\t"\ + "pcmpeqb %%mm7, %%mm4 \n\t"\ + "pcmpeqb %%mm7, %%mm5 \n\t"\ + "movq %%mm3, %%mm6 \n\t"\ + "pxor %%mm4, %%mm3 \n\t"\ + "pxor %%mm5, %%mm4 \n\t"\ + "pxor %%mm6, %%mm5 \n\t"\ + "por %%mm3, %%mm1 \n\t"\ + "por %%mm4, %%mm2 \n\t"\ + "por %%mm5, %%mm0 \n\t"\ + "pand %%mm2, %%mm0 \n\t"\ + "pand %%mm1, %%mm0 \n\t"\ + "movq %%mm0, " #b " \n\t" #define MEDIAN(a,b,c) REAL_MEDIAN(a,b,c) -MEDIAN((%0), (%%REGa), (%%REGa, %1)) +MEDIAN((%0) , (%%REGa) , (%%REGa, %1)) MEDIAN((%%REGa, %1), (%%REGa, %1, 2), (%0, %1, 4)) -MEDIAN((%0, %1, 4), (%%REGd), (%%REGd, %1)) +MEDIAN((%0, %1, 4) , (%%REGd) , (%%REGd, %1)) MEDIAN((%%REGd, %1), (%%REGd, %1, 2), (%0, %1, 8)) - : : "r" (src), "r" ((long)stride) - : "%"REG_a, "%"REG_d - ); + : : "r" (src), "r" ((long)stride) + : "%"REG_a, "%"REG_d + ); #endif //HAVE_MMX2 #else //HAVE_MMX - int x, y; - src+= 4*stride; - // FIXME - there should be a way to do a few columns in parallel like w/mmx - for(x=0; x<8; x++) - { - uint8_t *colsrc = src; - for (y=0; y<4; y++) - { - int a, b, c, d, e, f; - a = colsrc[0 ]; - b = colsrc[stride ]; - c = colsrc[stride*2]; - d = (a-b)>>31; - e = (b-c)>>31; - f = (c-a)>>31; - colsrc[stride ] = (a|(d^f)) & (b|(d^e)) & (c|(e^f)); - colsrc += stride*2; - } - src++; - } + int x, y; + src+= 4*stride; + // FIXME - there should be a way to do a few columns in parallel like w/mmx + for(x=0; x<8; x++) + { + uint8_t *colsrc = src; + for (y=0; y<4; y++) + { + int a, b, c, d, e, f; + a = colsrc[0 ]; + b = colsrc[stride ]; + c = colsrc[stride*2]; + d = (a-b)>>31; + e = (b-c)>>31; + f = (c-a)>>31; + colsrc[stride ] = (a|(d^f)) & (b|(d^e)) & (c|(e^f)); + colsrc += stride*2; + } + src++; + } #endif //HAVE_MMX } @@ -2072,84 +2072,84 @@ */ static inline void RENAME(transpose1)(uint8_t *dst1, uint8_t *dst2, uint8_t *src, int srcStride) { - asm( - "lea (%0, %1), %%"REG_a" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 - "movq (%0), %%mm0 \n\t" // 12345678 - "movq (%%"REG_a"), %%mm1 \n\t" // abcdefgh - "movq %%mm0, %%mm2 \n\t" // 12345678 - "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d - "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h - - "movq (%%"REG_a", %1), %%mm1 \n\t" - "movq (%%"REG_a", %1, 2), %%mm3 \n\t" - "movq %%mm1, %%mm4 \n\t" - "punpcklbw %%mm3, %%mm1 \n\t" - "punpckhbw %%mm3, %%mm4 \n\t" - - "movq %%mm0, %%mm3 \n\t" - "punpcklwd %%mm1, %%mm0 \n\t" - "punpckhwd %%mm1, %%mm3 \n\t" - "movq %%mm2, %%mm1 \n\t" - "punpcklwd %%mm4, %%mm2 \n\t" - "punpckhwd %%mm4, %%mm1 \n\t" - - "movd %%mm0, 128(%2) \n\t" - "psrlq $32, %%mm0 \n\t" - "movd %%mm0, 144(%2) \n\t" - "movd %%mm3, 160(%2) \n\t" - "psrlq $32, %%mm3 \n\t" - "movd %%mm3, 176(%2) \n\t" - "movd %%mm3, 48(%3) \n\t" - "movd %%mm2, 192(%2) \n\t" - "movd %%mm2, 64(%3) \n\t" - "psrlq $32, %%mm2 \n\t" - "movd %%mm2, 80(%3) \n\t" - "movd %%mm1, 96(%3) \n\t" - "psrlq $32, %%mm1 \n\t" - "movd %%mm1, 112(%3) \n\t" - - "lea (%%"REG_a", %1, 4), %%"REG_a" \n\t" - - "movq (%0, %1, 4), %%mm0 \n\t" // 12345678 - "movq (%%"REG_a"), %%mm1 \n\t" // abcdefgh - "movq %%mm0, %%mm2 \n\t" // 12345678 - "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d - "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h - - "movq (%%"REG_a", %1), %%mm1 \n\t" - "movq (%%"REG_a", %1, 2), %%mm3 \n\t" - "movq %%mm1, %%mm4 \n\t" - "punpcklbw %%mm3, %%mm1 \n\t" - "punpckhbw %%mm3, %%mm4 \n\t" - - "movq %%mm0, %%mm3 \n\t" - "punpcklwd %%mm1, %%mm0 \n\t" - "punpckhwd %%mm1, %%mm3 \n\t" - "movq %%mm2, %%mm1 \n\t" - "punpcklwd %%mm4, %%mm2 \n\t" - "punpckhwd %%mm4, %%mm1 \n\t" - - "movd %%mm0, 132(%2) \n\t" - "psrlq $32, %%mm0 \n\t" - "movd %%mm0, 148(%2) \n\t" - "movd %%mm3, 164(%2) \n\t" - "psrlq $32, %%mm3 \n\t" - "movd %%mm3, 180(%2) \n\t" - "movd %%mm3, 52(%3) \n\t" - "movd %%mm2, 196(%2) \n\t" - "movd %%mm2, 68(%3) \n\t" - "psrlq $32, %%mm2 \n\t" - "movd %%mm2, 84(%3) \n\t" - "movd %%mm1, 100(%3) \n\t" - "psrlq $32, %%mm1 \n\t" - "movd %%mm1, 116(%3) \n\t" - - - :: "r" (src), "r" ((long)srcStride), "r" (dst1), "r" (dst2) - : "%"REG_a - ); + asm( + "lea (%0, %1), %%"REG_a" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + "movq (%0), %%mm0 \n\t" // 12345678 + "movq (%%"REG_a"), %%mm1 \n\t" // abcdefgh + "movq %%mm0, %%mm2 \n\t" // 12345678 + "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d + "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h + + "movq (%%"REG_a", %1), %%mm1 \n\t" + "movq (%%"REG_a", %1, 2), %%mm3 \n\t" + "movq %%mm1, %%mm4 \n\t" + "punpcklbw %%mm3, %%mm1 \n\t" + "punpckhbw %%mm3, %%mm4 \n\t" + + "movq %%mm0, %%mm3 \n\t" + "punpcklwd %%mm1, %%mm0 \n\t" + "punpckhwd %%mm1, %%mm3 \n\t" + "movq %%mm2, %%mm1 \n\t" + "punpcklwd %%mm4, %%mm2 \n\t" + "punpckhwd %%mm4, %%mm1 \n\t" + + "movd %%mm0, 128(%2) \n\t" + "psrlq $32, %%mm0 \n\t" + "movd %%mm0, 144(%2) \n\t" + "movd %%mm3, 160(%2) \n\t" + "psrlq $32, %%mm3 \n\t" + "movd %%mm3, 176(%2) \n\t" + "movd %%mm3, 48(%3) \n\t" + "movd %%mm2, 192(%2) \n\t" + "movd %%mm2, 64(%3) \n\t" + "psrlq $32, %%mm2 \n\t" + "movd %%mm2, 80(%3) \n\t" + "movd %%mm1, 96(%3) \n\t" + "psrlq $32, %%mm1 \n\t" + "movd %%mm1, 112(%3) \n\t" + + "lea (%%"REG_a", %1, 4), %%"REG_a" \n\t" + + "movq (%0, %1, 4), %%mm0 \n\t" // 12345678 + "movq (%%"REG_a"), %%mm1 \n\t" // abcdefgh + "movq %%mm0, %%mm2 \n\t" // 12345678 + "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d + "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h + + "movq (%%"REG_a", %1), %%mm1 \n\t" + "movq (%%"REG_a", %1, 2), %%mm3 \n\t" + "movq %%mm1, %%mm4 \n\t" + "punpcklbw %%mm3, %%mm1 \n\t" + "punpckhbw %%mm3, %%mm4 \n\t" + + "movq %%mm0, %%mm3 \n\t" + "punpcklwd %%mm1, %%mm0 \n\t" + "punpckhwd %%mm1, %%mm3 \n\t" + "movq %%mm2, %%mm1 \n\t" + "punpcklwd %%mm4, %%mm2 \n\t" + "punpckhwd %%mm4, %%mm1 \n\t" + + "movd %%mm0, 132(%2) \n\t" + "psrlq $32, %%mm0 \n\t" + "movd %%mm0, 148(%2) \n\t" + "movd %%mm3, 164(%2) \n\t" + "psrlq $32, %%mm3 \n\t" + "movd %%mm3, 180(%2) \n\t" + "movd %%mm3, 52(%3) \n\t" + "movd %%mm2, 196(%2) \n\t" + "movd %%mm2, 68(%3) \n\t" + "psrlq $32, %%mm2 \n\t" + "movd %%mm2, 84(%3) \n\t" + "movd %%mm1, 100(%3) \n\t" + "psrlq $32, %%mm1 \n\t" + "movd %%mm1, 116(%3) \n\t" + + + :: "r" (src), "r" ((long)srcStride), "r" (dst1), "r" (dst2) + : "%"REG_a + ); } /** @@ -2157,414 +2157,414 @@ */ static inline void RENAME(transpose2)(uint8_t *dst, int dstStride, uint8_t *src) { - asm( - "lea (%0, %1), %%"REG_a" \n\t" - "lea (%%"REG_a",%1,4), %%"REG_d"\n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 - "movq (%2), %%mm0 \n\t" // 12345678 - "movq 16(%2), %%mm1 \n\t" // abcdefgh - "movq %%mm0, %%mm2 \n\t" // 12345678 - "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d - "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h - - "movq 32(%2), %%mm1 \n\t" - "movq 48(%2), %%mm3 \n\t" - "movq %%mm1, %%mm4 \n\t" - "punpcklbw %%mm3, %%mm1 \n\t" - "punpckhbw %%mm3, %%mm4 \n\t" - - "movq %%mm0, %%mm3 \n\t" - "punpcklwd %%mm1, %%mm0 \n\t" - "punpckhwd %%mm1, %%mm3 \n\t" - "movq %%mm2, %%mm1 \n\t" - "punpcklwd %%mm4, %%mm2 \n\t" - "punpckhwd %%mm4, %%mm1 \n\t" - - "movd %%mm0, (%0) \n\t" - "psrlq $32, %%mm0 \n\t" - "movd %%mm0, (%%"REG_a") \n\t" - "movd %%mm3, (%%"REG_a", %1) \n\t" - "psrlq $32, %%mm3 \n\t" - "movd %%mm3, (%%"REG_a", %1, 2) \n\t" - "movd %%mm2, (%0, %1, 4) \n\t" - "psrlq $32, %%mm2 \n\t" - "movd %%mm2, (%%"REG_d") \n\t" - "movd %%mm1, (%%"REG_d", %1) \n\t" - "psrlq $32, %%mm1 \n\t" - "movd %%mm1, (%%"REG_d", %1, 2) \n\t" - - - "movq 64(%2), %%mm0 \n\t" // 12345678 - "movq 80(%2), %%mm1 \n\t" // abcdefgh - "movq %%mm0, %%mm2 \n\t" // 12345678 - "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d - "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h - - "movq 96(%2), %%mm1 \n\t" - "movq 112(%2), %%mm3 \n\t" - "movq %%mm1, %%mm4 \n\t" - "punpcklbw %%mm3, %%mm1 \n\t" - "punpckhbw %%mm3, %%mm4 \n\t" - - "movq %%mm0, %%mm3 \n\t" - "punpcklwd %%mm1, %%mm0 \n\t" - "punpckhwd %%mm1, %%mm3 \n\t" - "movq %%mm2, %%mm1 \n\t" - "punpcklwd %%mm4, %%mm2 \n\t" - "punpckhwd %%mm4, %%mm1 \n\t" - - "movd %%mm0, 4(%0) \n\t" - "psrlq $32, %%mm0 \n\t" - "movd %%mm0, 4(%%"REG_a") \n\t" - "movd %%mm3, 4(%%"REG_a", %1) \n\t" - "psrlq $32, %%mm3 \n\t" - "movd %%mm3, 4(%%"REG_a", %1, 2) \n\t" - "movd %%mm2, 4(%0, %1, 4) \n\t" - "psrlq $32, %%mm2 \n\t" - "movd %%mm2, 4(%%"REG_d") \n\t" - "movd %%mm1, 4(%%"REG_d", %1) \n\t" - "psrlq $32, %%mm1 \n\t" - "movd %%mm1, 4(%%"REG_d", %1, 2) \n\t" - - :: "r" (dst), "r" ((long)dstStride), "r" (src) - : "%"REG_a, "%"REG_d - ); + asm( + "lea (%0, %1), %%"REG_a" \n\t" + "lea (%%"REG_a",%1,4), %%"REG_d" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 edx edx+%1 edx+2%1 %0+8%1 edx+4%1 + "movq (%2), %%mm0 \n\t" // 12345678 + "movq 16(%2), %%mm1 \n\t" // abcdefgh + "movq %%mm0, %%mm2 \n\t" // 12345678 + "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d + "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h + + "movq 32(%2), %%mm1 \n\t" + "movq 48(%2), %%mm3 \n\t" + "movq %%mm1, %%mm4 \n\t" + "punpcklbw %%mm3, %%mm1 \n\t" + "punpckhbw %%mm3, %%mm4 \n\t" + + "movq %%mm0, %%mm3 \n\t" + "punpcklwd %%mm1, %%mm0 \n\t" + "punpckhwd %%mm1, %%mm3 \n\t" + "movq %%mm2, %%mm1 \n\t" + "punpcklwd %%mm4, %%mm2 \n\t" + "punpckhwd %%mm4, %%mm1 \n\t" + + "movd %%mm0, (%0) \n\t" + "psrlq $32, %%mm0 \n\t" + "movd %%mm0, (%%"REG_a") \n\t" + "movd %%mm3, (%%"REG_a", %1) \n\t" + "psrlq $32, %%mm3 \n\t" + "movd %%mm3, (%%"REG_a", %1, 2) \n\t" + "movd %%mm2, (%0, %1, 4) \n\t" + "psrlq $32, %%mm2 \n\t" + "movd %%mm2, (%%"REG_d") \n\t" + "movd %%mm1, (%%"REG_d", %1) \n\t" + "psrlq $32, %%mm1 \n\t" + "movd %%mm1, (%%"REG_d", %1, 2) \n\t" + + + "movq 64(%2), %%mm0 \n\t" // 12345678 + "movq 80(%2), %%mm1 \n\t" // abcdefgh + "movq %%mm0, %%mm2 \n\t" // 12345678 + "punpcklbw %%mm1, %%mm0 \n\t" // 1a2b3c4d + "punpckhbw %%mm1, %%mm2 \n\t" // 5e6f7g8h + + "movq 96(%2), %%mm1 \n\t" + "movq 112(%2), %%mm3 \n\t" + "movq %%mm1, %%mm4 \n\t" + "punpcklbw %%mm3, %%mm1 \n\t" + "punpckhbw %%mm3, %%mm4 \n\t" + + "movq %%mm0, %%mm3 \n\t" + "punpcklwd %%mm1, %%mm0 \n\t" + "punpckhwd %%mm1, %%mm3 \n\t" + "movq %%mm2, %%mm1 \n\t" + "punpcklwd %%mm4, %%mm2 \n\t" + "punpckhwd %%mm4, %%mm1 \n\t" + + "movd %%mm0, 4(%0) \n\t" + "psrlq $32, %%mm0 \n\t" + "movd %%mm0, 4(%%"REG_a") \n\t" + "movd %%mm3, 4(%%"REG_a", %1) \n\t" + "psrlq $32, %%mm3 \n\t" + "movd %%mm3, 4(%%"REG_a", %1, 2) \n\t" + "movd %%mm2, 4(%0, %1, 4) \n\t" + "psrlq $32, %%mm2 \n\t" + "movd %%mm2, 4(%%"REG_d") \n\t" + "movd %%mm1, 4(%%"REG_d", %1) \n\t" + "psrlq $32, %%mm1 \n\t" + "movd %%mm1, 4(%%"REG_d", %1, 2) \n\t" + + :: "r" (dst), "r" ((long)dstStride), "r" (src) + : "%"REG_a, "%"REG_d + ); } #endif //HAVE_MMX //static long test=0; #ifndef HAVE_ALTIVEC static inline void RENAME(tempNoiseReducer)(uint8_t *src, int stride, - uint8_t *tempBlured, uint32_t *tempBluredPast, int *maxNoise) + uint8_t *tempBlured, uint32_t *tempBluredPast, int *maxNoise) { - // to save a register (FIXME do this outside of the loops) - tempBluredPast[127]= maxNoise[0]; - tempBluredPast[128]= maxNoise[1]; - tempBluredPast[129]= maxNoise[2]; + // to save a register (FIXME do this outside of the loops) + tempBluredPast[127]= maxNoise[0]; + tempBluredPast[128]= maxNoise[1]; + tempBluredPast[129]= maxNoise[2]; #define FAST_L2_DIFF //#define L1_DIFF //u should change the thresholds too if u try that one #if defined (HAVE_MMX2) || defined (HAVE_3DNOW) - asm volatile( - "lea (%2, %2, 2), %%"REG_a" \n\t" // 3*stride - "lea (%2, %2, 4), %%"REG_d" \n\t" // 5*stride - "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride -// 0 1 2 3 4 5 6 7 8 9 -// %x %x+%2 %x+2%2 %x+eax %x+4%2 %x+edx %x+2eax %x+ecx %x+8%2 + asm volatile( + "lea (%2, %2, 2), %%"REG_a" \n\t" // 3*stride + "lea (%2, %2, 4), %%"REG_d" \n\t" // 5*stride + "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride +// 0 1 2 3 4 5 6 7 8 9 +// %x %x+%2 %x+2%2 %x+eax %x+4%2 %x+edx %x+2eax %x+ecx %x+8%2 //FIXME reorder? #ifdef L1_DIFF //needs mmx2 - "movq (%0), %%mm0 \n\t" // L0 - "psadbw (%1), %%mm0 \n\t" // |L0-R0| - "movq (%0, %2), %%mm1 \n\t" // L1 - "psadbw (%1, %2), %%mm1 \n\t" // |L1-R1| - "movq (%0, %2, 2), %%mm2 \n\t" // L2 - "psadbw (%1, %2, 2), %%mm2 \n\t" // |L2-R2| - "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 - "psadbw (%1, %%"REG_a"), %%mm3 \n\t" // |L3-R3| - - "movq (%0, %2, 4), %%mm4 \n\t" // L4 - "paddw %%mm1, %%mm0 \n\t" - "psadbw (%1, %2, 4), %%mm4 \n\t" // |L4-R4| - "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 - "paddw %%mm2, %%mm0 \n\t" - "psadbw (%1, %%"REG_d"), %%mm5 \n\t" // |L5-R5| - "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 - "paddw %%mm3, %%mm0 \n\t" - "psadbw (%1, %%"REG_a", 2), %%mm6 \n\t" // |L6-R6| - "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 - "paddw %%mm4, %%mm0 \n\t" - "psadbw (%1, %%"REG_c"), %%mm7 \n\t" // |L7-R7| - "paddw %%mm5, %%mm6 \n\t" - "paddw %%mm7, %%mm6 \n\t" - "paddw %%mm6, %%mm0 \n\t" + "movq (%0), %%mm0 \n\t" // L0 + "psadbw (%1), %%mm0 \n\t" // |L0-R0| + "movq (%0, %2), %%mm1 \n\t" // L1 + "psadbw (%1, %2), %%mm1 \n\t" // |L1-R1| + "movq (%0, %2, 2), %%mm2 \n\t" // L2 + "psadbw (%1, %2, 2), %%mm2 \n\t" // |L2-R2| + "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 + "psadbw (%1, %%"REG_a"), %%mm3 \n\t" // |L3-R3| + + "movq (%0, %2, 4), %%mm4 \n\t" // L4 + "paddw %%mm1, %%mm0 \n\t" + "psadbw (%1, %2, 4), %%mm4 \n\t" // |L4-R4| + "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 + "paddw %%mm2, %%mm0 \n\t" + "psadbw (%1, %%"REG_d"), %%mm5 \n\t" // |L5-R5| + "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 + "paddw %%mm3, %%mm0 \n\t" + "psadbw (%1, %%"REG_a", 2), %%mm6 \n\t" // |L6-R6| + "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 + "paddw %%mm4, %%mm0 \n\t" + "psadbw (%1, %%"REG_c"), %%mm7 \n\t" // |L7-R7| + "paddw %%mm5, %%mm6 \n\t" + "paddw %%mm7, %%mm6 \n\t" + "paddw %%mm6, %%mm0 \n\t" #else //L1_DIFF #if defined (FAST_L2_DIFF) - "pcmpeqb %%mm7, %%mm7 \n\t" - "movq "MANGLE(b80)", %%mm6 \n\t" - "pxor %%mm0, %%mm0 \n\t" + "pcmpeqb %%mm7, %%mm7 \n\t" + "movq "MANGLE(b80)", %%mm6 \n\t" + "pxor %%mm0, %%mm0 \n\t" #define REAL_L2_DIFF_CORE(a, b)\ - "movq " #a ", %%mm5 \n\t"\ - "movq " #b ", %%mm2 \n\t"\ - "pxor %%mm7, %%mm2 \n\t"\ - PAVGB(%%mm2, %%mm5)\ - "paddb %%mm6, %%mm5 \n\t"\ - "movq %%mm5, %%mm2 \n\t"\ - "psllw $8, %%mm5 \n\t"\ - "pmaddwd %%mm5, %%mm5 \n\t"\ - "pmaddwd %%mm2, %%mm2 \n\t"\ - "paddd %%mm2, %%mm5 \n\t"\ - "psrld $14, %%mm5 \n\t"\ - "paddd %%mm5, %%mm0 \n\t" + "movq " #a ", %%mm5 \n\t"\ + "movq " #b ", %%mm2 \n\t"\ + "pxor %%mm7, %%mm2 \n\t"\ + PAVGB(%%mm2, %%mm5)\ + "paddb %%mm6, %%mm5 \n\t"\ + "movq %%mm5, %%mm2 \n\t"\ + "psllw $8, %%mm5 \n\t"\ + "pmaddwd %%mm5, %%mm5 \n\t"\ + "pmaddwd %%mm2, %%mm2 \n\t"\ + "paddd %%mm2, %%mm5 \n\t"\ + "psrld $14, %%mm5 \n\t"\ + "paddd %%mm5, %%mm0 \n\t" #else //defined (FAST_L2_DIFF) - "pxor %%mm7, %%mm7 \n\t" - "pxor %%mm0, %%mm0 \n\t" + "pxor %%mm7, %%mm7 \n\t" + "pxor %%mm0, %%mm0 \n\t" #define REAL_L2_DIFF_CORE(a, b)\ - "movq " #a ", %%mm5 \n\t"\ - "movq " #b ", %%mm2 \n\t"\ - "movq %%mm5, %%mm1 \n\t"\ - "movq %%mm2, %%mm3 \n\t"\ - "punpcklbw %%mm7, %%mm5 \n\t"\ - "punpckhbw %%mm7, %%mm1 \n\t"\ - "punpcklbw %%mm7, %%mm2 \n\t"\ - "punpckhbw %%mm7, %%mm3 \n\t"\ - "psubw %%mm2, %%mm5 \n\t"\ - "psubw %%mm3, %%mm1 \n\t"\ - "pmaddwd %%mm5, %%mm5 \n\t"\ - "pmaddwd %%mm1, %%mm1 \n\t"\ - "paddd %%mm1, %%mm5 \n\t"\ - "paddd %%mm5, %%mm0 \n\t" + "movq " #a ", %%mm5 \n\t"\ + "movq " #b ", %%mm2 \n\t"\ + "movq %%mm5, %%mm1 \n\t"\ + "movq %%mm2, %%mm3 \n\t"\ + "punpcklbw %%mm7, %%mm5 \n\t"\ + "punpckhbw %%mm7, %%mm1 \n\t"\ + "punpcklbw %%mm7, %%mm2 \n\t"\ + "punpckhbw %%mm7, %%mm3 \n\t"\ + "psubw %%mm2, %%mm5 \n\t"\ + "psubw %%mm3, %%mm1 \n\t"\ + "pmaddwd %%mm5, %%mm5 \n\t"\ + "pmaddwd %%mm1, %%mm1 \n\t"\ + "paddd %%mm1, %%mm5 \n\t"\ + "paddd %%mm5, %%mm0 \n\t" #endif //defined (FAST_L2_DIFF) #define L2_DIFF_CORE(a, b) REAL_L2_DIFF_CORE(a, b) -L2_DIFF_CORE((%0), (%1)) -L2_DIFF_CORE((%0, %2), (%1, %2)) -L2_DIFF_CORE((%0, %2, 2), (%1, %2, 2)) -L2_DIFF_CORE((%0, %%REGa), (%1, %%REGa)) -L2_DIFF_CORE((%0, %2, 4), (%1, %2, 4)) -L2_DIFF_CORE((%0, %%REGd), (%1, %%REGd)) +L2_DIFF_CORE((%0) , (%1)) +L2_DIFF_CORE((%0, %2) , (%1, %2)) +L2_DIFF_CORE((%0, %2, 2) , (%1, %2, 2)) +L2_DIFF_CORE((%0, %%REGa) , (%1, %%REGa)) +L2_DIFF_CORE((%0, %2, 4) , (%1, %2, 4)) +L2_DIFF_CORE((%0, %%REGd) , (%1, %%REGd)) L2_DIFF_CORE((%0, %%REGa,2), (%1, %%REGa,2)) -L2_DIFF_CORE((%0, %%REGc), (%1, %%REGc)) +L2_DIFF_CORE((%0, %%REGc) , (%1, %%REGc)) #endif //L1_DIFF - "movq %%mm0, %%mm4 \n\t" - "psrlq $32, %%mm0 \n\t" - "paddd %%mm0, %%mm4 \n\t" - "movd %%mm4, %%ecx \n\t" - "shll $2, %%ecx \n\t" - "mov %3, %%"REG_d" \n\t" - "addl -4(%%"REG_d"), %%ecx \n\t" - "addl 4(%%"REG_d"), %%ecx \n\t" - "addl -1024(%%"REG_d"), %%ecx \n\t" - "addl $4, %%ecx \n\t" - "addl 1024(%%"REG_d"), %%ecx \n\t" - "shrl $3, %%ecx \n\t" - "movl %%ecx, (%%"REG_d") \n\t" - -// "mov %3, %%"REG_c" \n\t" -// "mov %%"REG_c", test \n\t" -// "jmp 4f \n\t" - "cmpl 512(%%"REG_d"), %%ecx \n\t" - " jb 2f \n\t" - "cmpl 516(%%"REG_d"), %%ecx \n\t" - " jb 1f \n\t" - - "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride - "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride - "movq (%0), %%mm0 \n\t" // L0 - "movq (%0, %2), %%mm1 \n\t" // L1 - "movq (%0, %2, 2), %%mm2 \n\t" // L2 - "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 - "movq (%0, %2, 4), %%mm4 \n\t" // L4 - "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 - "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 - "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 - "movq %%mm0, (%1) \n\t" // L0 - "movq %%mm1, (%1, %2) \n\t" // L1 - "movq %%mm2, (%1, %2, 2) \n\t" // L2 - "movq %%mm3, (%1, %%"REG_a") \n\t" // L3 - "movq %%mm4, (%1, %2, 4) \n\t" // L4 - "movq %%mm5, (%1, %%"REG_d") \n\t" // L5 - "movq %%mm6, (%1, %%"REG_a", 2) \n\t" // L6 - "movq %%mm7, (%1, %%"REG_c") \n\t" // L7 - "jmp 4f \n\t" - - "1: \n\t" - "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride - "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride - "movq (%0), %%mm0 \n\t" // L0 - PAVGB((%1), %%mm0) // L0 - "movq (%0, %2), %%mm1 \n\t" // L1 - PAVGB((%1, %2), %%mm1) // L1 - "movq (%0, %2, 2), %%mm2 \n\t" // L2 - PAVGB((%1, %2, 2), %%mm2) // L2 - "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 - PAVGB((%1, %%REGa), %%mm3) // L3 - "movq (%0, %2, 4), %%mm4 \n\t" // L4 - PAVGB((%1, %2, 4), %%mm4) // L4 - "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 - PAVGB((%1, %%REGd), %%mm5) // L5 - "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 - PAVGB((%1, %%REGa, 2), %%mm6) // L6 - "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 - PAVGB((%1, %%REGc), %%mm7) // L7 - "movq %%mm0, (%1) \n\t" // R0 - "movq %%mm1, (%1, %2) \n\t" // R1 - "movq %%mm2, (%1, %2, 2) \n\t" // R2 - "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 - "movq %%mm4, (%1, %2, 4) \n\t" // R4 - "movq %%mm5, (%1, %%"REG_d") \n\t" // R5 - "movq %%mm6, (%1, %%"REG_a", 2) \n\t" // R6 - "movq %%mm7, (%1, %%"REG_c") \n\t" // R7 - "movq %%mm0, (%0) \n\t" // L0 - "movq %%mm1, (%0, %2) \n\t" // L1 - "movq %%mm2, (%0, %2, 2) \n\t" // L2 - "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 - "movq %%mm4, (%0, %2, 4) \n\t" // L4 - "movq %%mm5, (%0, %%"REG_d") \n\t" // L5 - "movq %%mm6, (%0, %%"REG_a", 2) \n\t" // L6 - "movq %%mm7, (%0, %%"REG_c") \n\t" // L7 - "jmp 4f \n\t" - - "2: \n\t" - "cmpl 508(%%"REG_d"), %%ecx \n\t" - " jb 3f \n\t" - - "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride - "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride - "movq (%0), %%mm0 \n\t" // L0 - "movq (%0, %2), %%mm1 \n\t" // L1 - "movq (%0, %2, 2), %%mm2 \n\t" // L2 - "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 - "movq (%1), %%mm4 \n\t" // R0 - "movq (%1, %2), %%mm5 \n\t" // R1 - "movq (%1, %2, 2), %%mm6 \n\t" // R2 - "movq (%1, %%"REG_a"), %%mm7 \n\t" // R3 - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - "movq %%mm0, (%1) \n\t" // R0 - "movq %%mm1, (%1, %2) \n\t" // R1 - "movq %%mm2, (%1, %2, 2) \n\t" // R2 - "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 - "movq %%mm0, (%0) \n\t" // L0 - "movq %%mm1, (%0, %2) \n\t" // L1 - "movq %%mm2, (%0, %2, 2) \n\t" // L2 - "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 - - "movq (%0, %2, 4), %%mm0 \n\t" // L4 - "movq (%0, %%"REG_d"), %%mm1 \n\t" // L5 - "movq (%0, %%"REG_a", 2), %%mm2 \n\t" // L6 - "movq (%0, %%"REG_c"), %%mm3 \n\t" // L7 - "movq (%1, %2, 4), %%mm4 \n\t" // R4 - "movq (%1, %%"REG_d"), %%mm5 \n\t" // R5 - "movq (%1, %%"REG_a", 2), %%mm6 \n\t" // R6 - "movq (%1, %%"REG_c"), %%mm7 \n\t" // R7 - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - "movq %%mm0, (%1, %2, 4) \n\t" // R4 - "movq %%mm1, (%1, %%"REG_d") \n\t" // R5 - "movq %%mm2, (%1, %%"REG_a", 2) \n\t" // R6 - "movq %%mm3, (%1, %%"REG_c") \n\t" // R7 - "movq %%mm0, (%0, %2, 4) \n\t" // L4 - "movq %%mm1, (%0, %%"REG_d") \n\t" // L5 - "movq %%mm2, (%0, %%"REG_a", 2) \n\t" // L6 - "movq %%mm3, (%0, %%"REG_c") \n\t" // L7 - "jmp 4f \n\t" - - "3: \n\t" - "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride - "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride - "movq (%0), %%mm0 \n\t" // L0 - "movq (%0, %2), %%mm1 \n\t" // L1 - "movq (%0, %2, 2), %%mm2 \n\t" // L2 - "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 - "movq (%1), %%mm4 \n\t" // R0 - "movq (%1, %2), %%mm5 \n\t" // R1 - "movq (%1, %2, 2), %%mm6 \n\t" // R2 - "movq (%1, %%"REG_a"), %%mm7 \n\t" // R3 - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - "movq %%mm0, (%1) \n\t" // R0 - "movq %%mm1, (%1, %2) \n\t" // R1 - "movq %%mm2, (%1, %2, 2) \n\t" // R2 - "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 - "movq %%mm0, (%0) \n\t" // L0 - "movq %%mm1, (%0, %2) \n\t" // L1 - "movq %%mm2, (%0, %2, 2) \n\t" // L2 - "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 - - "movq (%0, %2, 4), %%mm0 \n\t" // L4 - "movq (%0, %%"REG_d"), %%mm1 \n\t" // L5 - "movq (%0, %%"REG_a", 2), %%mm2 \n\t" // L6 - "movq (%0, %%"REG_c"), %%mm3 \n\t" // L7 - "movq (%1, %2, 4), %%mm4 \n\t" // R4 - "movq (%1, %%"REG_d"), %%mm5 \n\t" // R5 - "movq (%1, %%"REG_a", 2), %%mm6 \n\t" // R6 - "movq (%1, %%"REG_c"), %%mm7 \n\t" // R7 - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - PAVGB(%%mm4, %%mm0) - PAVGB(%%mm5, %%mm1) - PAVGB(%%mm6, %%mm2) - PAVGB(%%mm7, %%mm3) - "movq %%mm0, (%1, %2, 4) \n\t" // R4 - "movq %%mm1, (%1, %%"REG_d") \n\t" // R5 - "movq %%mm2, (%1, %%"REG_a", 2) \n\t" // R6 - "movq %%mm3, (%1, %%"REG_c") \n\t" // R7 - "movq %%mm0, (%0, %2, 4) \n\t" // L4 - "movq %%mm1, (%0, %%"REG_d") \n\t" // L5 - "movq %%mm2, (%0, %%"REG_a", 2) \n\t" // L6 - "movq %%mm3, (%0, %%"REG_c") \n\t" // L7 - - "4: \n\t" - - :: "r" (src), "r" (tempBlured), "r"((long)stride), "m" (tempBluredPast) - : "%"REG_a, "%"REG_d, "%"REG_c, "memory" - ); + "movq %%mm0, %%mm4 \n\t" + "psrlq $32, %%mm0 \n\t" + "paddd %%mm0, %%mm4 \n\t" + "movd %%mm4, %%ecx \n\t" + "shll $2, %%ecx \n\t" + "mov %3, %%"REG_d" \n\t" + "addl -4(%%"REG_d"), %%ecx \n\t" + "addl 4(%%"REG_d"), %%ecx \n\t" + "addl -1024(%%"REG_d"), %%ecx \n\t" + "addl $4, %%ecx \n\t" + "addl 1024(%%"REG_d"), %%ecx \n\t" + "shrl $3, %%ecx \n\t" + "movl %%ecx, (%%"REG_d") \n\t" + +// "mov %3, %%"REG_c" \n\t" +// "mov %%"REG_c", test \n\t" +// "jmp 4f \n\t" + "cmpl 512(%%"REG_d"), %%ecx \n\t" + " jb 2f \n\t" + "cmpl 516(%%"REG_d"), %%ecx \n\t" + " jb 1f \n\t" + + "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride + "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride + "movq (%0), %%mm0 \n\t" // L0 + "movq (%0, %2), %%mm1 \n\t" // L1 + "movq (%0, %2, 2), %%mm2 \n\t" // L2 + "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 + "movq (%0, %2, 4), %%mm4 \n\t" // L4 + "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 + "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 + "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 + "movq %%mm0, (%1) \n\t" // L0 + "movq %%mm1, (%1, %2) \n\t" // L1 + "movq %%mm2, (%1, %2, 2) \n\t" // L2 + "movq %%mm3, (%1, %%"REG_a") \n\t" // L3 + "movq %%mm4, (%1, %2, 4) \n\t" // L4 + "movq %%mm5, (%1, %%"REG_d") \n\t" // L5 + "movq %%mm6, (%1, %%"REG_a", 2) \n\t" // L6 + "movq %%mm7, (%1, %%"REG_c") \n\t" // L7 + "jmp 4f \n\t" + + "1: \n\t" + "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride + "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride + "movq (%0), %%mm0 \n\t" // L0 + PAVGB((%1), %%mm0) // L0 + "movq (%0, %2), %%mm1 \n\t" // L1 + PAVGB((%1, %2), %%mm1) // L1 + "movq (%0, %2, 2), %%mm2 \n\t" // L2 + PAVGB((%1, %2, 2), %%mm2) // L2 + "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 + PAVGB((%1, %%REGa), %%mm3) // L3 + "movq (%0, %2, 4), %%mm4 \n\t" // L4 + PAVGB((%1, %2, 4), %%mm4) // L4 + "movq (%0, %%"REG_d"), %%mm5 \n\t" // L5 + PAVGB((%1, %%REGd), %%mm5) // L5 + "movq (%0, %%"REG_a", 2), %%mm6 \n\t" // L6 + PAVGB((%1, %%REGa, 2), %%mm6) // L6 + "movq (%0, %%"REG_c"), %%mm7 \n\t" // L7 + PAVGB((%1, %%REGc), %%mm7) // L7 + "movq %%mm0, (%1) \n\t" // R0 + "movq %%mm1, (%1, %2) \n\t" // R1 + "movq %%mm2, (%1, %2, 2) \n\t" // R2 + "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 + "movq %%mm4, (%1, %2, 4) \n\t" // R4 + "movq %%mm5, (%1, %%"REG_d") \n\t" // R5 + "movq %%mm6, (%1, %%"REG_a", 2) \n\t" // R6 + "movq %%mm7, (%1, %%"REG_c") \n\t" // R7 + "movq %%mm0, (%0) \n\t" // L0 + "movq %%mm1, (%0, %2) \n\t" // L1 + "movq %%mm2, (%0, %2, 2) \n\t" // L2 + "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 + "movq %%mm4, (%0, %2, 4) \n\t" // L4 + "movq %%mm5, (%0, %%"REG_d") \n\t" // L5 + "movq %%mm6, (%0, %%"REG_a", 2) \n\t" // L6 + "movq %%mm7, (%0, %%"REG_c") \n\t" // L7 + "jmp 4f \n\t" + + "2: \n\t" + "cmpl 508(%%"REG_d"), %%ecx \n\t" + " jb 3f \n\t" + + "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride + "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride + "movq (%0), %%mm0 \n\t" // L0 + "movq (%0, %2), %%mm1 \n\t" // L1 + "movq (%0, %2, 2), %%mm2 \n\t" // L2 + "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 + "movq (%1), %%mm4 \n\t" // R0 + "movq (%1, %2), %%mm5 \n\t" // R1 + "movq (%1, %2, 2), %%mm6 \n\t" // R2 + "movq (%1, %%"REG_a"), %%mm7 \n\t" // R3 + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + "movq %%mm0, (%1) \n\t" // R0 + "movq %%mm1, (%1, %2) \n\t" // R1 + "movq %%mm2, (%1, %2, 2) \n\t" // R2 + "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 + "movq %%mm0, (%0) \n\t" // L0 + "movq %%mm1, (%0, %2) \n\t" // L1 + "movq %%mm2, (%0, %2, 2) \n\t" // L2 + "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 + + "movq (%0, %2, 4), %%mm0 \n\t" // L4 + "movq (%0, %%"REG_d"), %%mm1 \n\t" // L5 + "movq (%0, %%"REG_a", 2), %%mm2 \n\t" // L6 + "movq (%0, %%"REG_c"), %%mm3 \n\t" // L7 + "movq (%1, %2, 4), %%mm4 \n\t" // R4 + "movq (%1, %%"REG_d"), %%mm5 \n\t" // R5 + "movq (%1, %%"REG_a", 2), %%mm6 \n\t" // R6 + "movq (%1, %%"REG_c"), %%mm7 \n\t" // R7 + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + "movq %%mm0, (%1, %2, 4) \n\t" // R4 + "movq %%mm1, (%1, %%"REG_d") \n\t" // R5 + "movq %%mm2, (%1, %%"REG_a", 2) \n\t" // R6 + "movq %%mm3, (%1, %%"REG_c") \n\t" // R7 + "movq %%mm0, (%0, %2, 4) \n\t" // L4 + "movq %%mm1, (%0, %%"REG_d") \n\t" // L5 + "movq %%mm2, (%0, %%"REG_a", 2) \n\t" // L6 + "movq %%mm3, (%0, %%"REG_c") \n\t" // L7 + "jmp 4f \n\t" + + "3: \n\t" + "lea (%%"REG_a", %2, 2), %%"REG_d" \n\t" // 5*stride + "lea (%%"REG_d", %2, 2), %%"REG_c" \n\t" // 7*stride + "movq (%0), %%mm0 \n\t" // L0 + "movq (%0, %2), %%mm1 \n\t" // L1 + "movq (%0, %2, 2), %%mm2 \n\t" // L2 + "movq (%0, %%"REG_a"), %%mm3 \n\t" // L3 + "movq (%1), %%mm4 \n\t" // R0 + "movq (%1, %2), %%mm5 \n\t" // R1 + "movq (%1, %2, 2), %%mm6 \n\t" // R2 + "movq (%1, %%"REG_a"), %%mm7 \n\t" // R3 + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + "movq %%mm0, (%1) \n\t" // R0 + "movq %%mm1, (%1, %2) \n\t" // R1 + "movq %%mm2, (%1, %2, 2) \n\t" // R2 + "movq %%mm3, (%1, %%"REG_a") \n\t" // R3 + "movq %%mm0, (%0) \n\t" // L0 + "movq %%mm1, (%0, %2) \n\t" // L1 + "movq %%mm2, (%0, %2, 2) \n\t" // L2 + "movq %%mm3, (%0, %%"REG_a") \n\t" // L3 + + "movq (%0, %2, 4), %%mm0 \n\t" // L4 + "movq (%0, %%"REG_d"), %%mm1 \n\t" // L5 + "movq (%0, %%"REG_a", 2), %%mm2 \n\t" // L6 + "movq (%0, %%"REG_c"), %%mm3 \n\t" // L7 + "movq (%1, %2, 4), %%mm4 \n\t" // R4 + "movq (%1, %%"REG_d"), %%mm5 \n\t" // R5 + "movq (%1, %%"REG_a", 2), %%mm6 \n\t" // R6 + "movq (%1, %%"REG_c"), %%mm7 \n\t" // R7 + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + PAVGB(%%mm4, %%mm0) + PAVGB(%%mm5, %%mm1) + PAVGB(%%mm6, %%mm2) + PAVGB(%%mm7, %%mm3) + "movq %%mm0, (%1, %2, 4) \n\t" // R4 + "movq %%mm1, (%1, %%"REG_d") \n\t" // R5 + "movq %%mm2, (%1, %%"REG_a", 2) \n\t" // R6 + "movq %%mm3, (%1, %%"REG_c") \n\t" // R7 + "movq %%mm0, (%0, %2, 4) \n\t" // L4 + "movq %%mm1, (%0, %%"REG_d") \n\t" // L5 + "movq %%mm2, (%0, %%"REG_a", 2) \n\t" // L6 + "movq %%mm3, (%0, %%"REG_c") \n\t" // L7 + + "4: \n\t" + + :: "r" (src), "r" (tempBlured), "r"((long)stride), "m" (tempBluredPast) + : "%"REG_a, "%"REG_d, "%"REG_c, "memory" + ); //printf("%d\n", test); #else //defined (HAVE_MMX2) || defined (HAVE_3DNOW) { - int y; - int d=0; -// int sysd=0; - int i; - - for(y=0; y<8; y++) - { - int x; - for(x=0; x<8; x++) - { - int ref= tempBlured[ x + y*stride ]; - int cur= src[ x + y*stride ]; - int d1=ref - cur; -// if(x==0 || x==7) d1+= d1>>1; -// if(y==0 || y==7) d1+= d1>>1; -// d+= ABS(d1); - d+= d1*d1; -// sysd+= d1; - } - } - i=d; - d= ( - 4*d - +(*(tempBluredPast-256)) - +(*(tempBluredPast-1))+ (*(tempBluredPast+1)) - +(*(tempBluredPast+256)) - +4)>>3; - *tempBluredPast=i; -// ((*tempBluredPast)*3 + d + 2)>>2; + int y; + int d=0; +// int sysd=0; + int i; + + for(y=0; y<8; y++) + { + int x; + for(x=0; x<8; x++) + { + int ref= tempBlured[ x + y*stride ]; + int cur= src[ x + y*stride ]; + int d1=ref - cur; +// if(x==0 || x==7) d1+= d1>>1; +// if(y==0 || y==7) d1+= d1>>1; +// d+= ABS(d1); + d+= d1*d1; +// sysd+= d1; + } + } + i=d; + d= ( + 4*d + +(*(tempBluredPast-256)) + +(*(tempBluredPast-1))+ (*(tempBluredPast+1)) + +(*(tempBluredPast+256)) + +4)>>3; + *tempBluredPast=i; +// ((*tempBluredPast)*3 + d + 2)>>2; //printf("%d %d %d\n", maxNoise[0], maxNoise[1], maxNoise[2]); /* @@ -2574,68 +2574,68 @@ 64 48 36 27 20 15 11 (33) (approx) 64 56 49 43 37 33 29 (200) (approx) */ - if(d > maxNoise[1]) - { - if(d < maxNoise[2]) - { - for(y=0; y<8; y++) - { - int x; - for(x=0; x<8; x++) - { - int ref= tempBlured[ x + y*stride ]; - int cur= src[ x + y*stride ]; - tempBlured[ x + y*stride ]= - src[ x + y*stride ]= - (ref + cur + 1)>>1; - } - } - } - else - { - for(y=0; y<8; y++) - { - int x; - for(x=0; x<8; x++) - { - tempBlured[ x + y*stride ]= src[ x + y*stride ]; - } - } - } - } - else - { - if(d < maxNoise[0]) - { - for(y=0; y<8; y++) - { - int x; - for(x=0; x<8; x++) - { - int ref= tempBlured[ x + y*stride ]; - int cur= src[ x + y*stride ]; - tempBlured[ x + y*stride ]= - src[ x + y*stride ]= - (ref*7 + cur + 4)>>3; - } - } - } - else - { - for(y=0; y<8; y++) - { - int x; - for(x=0; x<8; x++) - { - int ref= tempBlured[ x + y*stride ]; - int cur= src[ x + y*stride ]; - tempBlured[ x + y*stride ]= - src[ x + y*stride ]= - (ref*3 + cur + 2)>>2; - } - } - } - } + if(d > maxNoise[1]) + { + if(d < maxNoise[2]) + { + for(y=0; y<8; y++) + { + int x; + for(x=0; x<8; x++) + { + int ref= tempBlured[ x + y*stride ]; + int cur= src[ x + y*stride ]; + tempBlured[ x + y*stride ]= + src[ x + y*stride ]= + (ref + cur + 1)>>1; + } + } + } + else + { + for(y=0; y<8; y++) + { + int x; + for(x=0; x<8; x++) + { + tempBlured[ x + y*stride ]= src[ x + y*stride ]; + } + } + } + } + else + { + if(d < maxNoise[0]) + { + for(y=0; y<8; y++) + { + int x; + for(x=0; x<8; x++) + { + int ref= tempBlured[ x + y*stride ]; + int cur= src[ x + y*stride ]; + tempBlured[ x + y*stride ]= + src[ x + y*stride ]= + (ref*7 + cur + 4)>>3; + } + } + } + else + { + for(y=0; y<8; y++) + { + int x; + for(x=0; x<8; x++) + { + int ref= tempBlured[ x + y*stride ]; + int cur= src[ x + y*stride ]; + tempBlured[ x + y*stride ]= + src[ x + y*stride ]= + (ref*3 + cur + 2)>>2; + } + } + } + } } #endif //defined (HAVE_MMX2) || defined (HAVE_3DNOW) } @@ -2646,531 +2646,531 @@ * accurate deblock filter */ static always_inline void RENAME(do_a_deblock)(uint8_t *src, int step, int stride, PPContext *c){ - int64_t dc_mask, eq_mask, both_masks; - int64_t sums[10*8*2]; - src+= step*3; // src points to begin of the 8x8 Block + int64_t dc_mask, eq_mask, both_masks; + int64_t sums[10*8*2]; + src+= step*3; // src points to begin of the 8x8 Block //START_TIMER asm volatile( - "movq %0, %%mm7 \n\t" - "movq %1, %%mm6 \n\t" + "movq %0, %%mm7 \n\t" + "movq %1, %%mm6 \n\t" : : "m" (c->mmxDcOffset[c->nonBQP]), "m" (c->mmxDcThreshold[c->nonBQP]) ); asm volatile( - "lea (%2, %3), %%"REG_a" \n\t" -// 0 1 2 3 4 5 6 7 8 9 -// %1 eax eax+%2 eax+2%2 %1+4%2 ecx ecx+%2 ecx+2%2 %1+8%2 ecx+4%2 - - "movq (%2), %%mm0 \n\t" - "movq (%%"REG_a"), %%mm1 \n\t" - "movq %%mm1, %%mm3 \n\t" - "movq %%mm1, %%mm4 \n\t" - "psubb %%mm1, %%mm0 \n\t" // mm0 = differnece - "paddb %%mm7, %%mm0 \n\t" - "pcmpgtb %%mm6, %%mm0 \n\t" - - "movq (%%"REG_a",%3), %%mm2 \n\t" + "lea (%2, %3), %%"REG_a" \n\t" +// 0 1 2 3 4 5 6 7 8 9 +// %1 eax eax+%2 eax+2%2 %1+4%2 ecx ecx+%2 ecx+2%2 %1+8%2 ecx+4%2 + + "movq (%2), %%mm0 \n\t" + "movq (%%"REG_a"), %%mm1 \n\t" + "movq %%mm1, %%mm3 \n\t" + "movq %%mm1, %%mm4 \n\t" + "psubb %%mm1, %%mm0 \n\t" // mm0 = differnece + "paddb %%mm7, %%mm0 \n\t" + "pcmpgtb %%mm6, %%mm0 \n\t" + + "movq (%%"REG_a",%3), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a", %3, 2), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a", %3, 2), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - - "lea (%%"REG_a", %3, 4), %%"REG_a" \n\t" - - "movq (%2, %3, 4), %%mm2 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + + "lea (%%"REG_a", %3, 4), %%"REG_a" \n\t" + + "movq (%2, %3, 4), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a"), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a"), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - - "movq (%%"REG_a", %3), %%mm2 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + + "movq (%%"REG_a", %3), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a", %3, 2), %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a", %3, 2), %%mm1 \n\t" PMAXUB(%%mm1, %%mm4) PMINUB(%%mm1, %%mm3, %%mm5) - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - - "movq (%2, %3, 8), %%mm2 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + + "movq (%2, %3, 8), %%mm2 \n\t" PMAXUB(%%mm2, %%mm4) PMINUB(%%mm2, %%mm3, %%mm5) - "psubb %%mm2, %%mm1 \n\t" - "paddb %%mm7, %%mm1 \n\t" - "pcmpgtb %%mm6, %%mm1 \n\t" - "paddb %%mm1, %%mm0 \n\t" - - "movq (%%"REG_a", %3, 4), %%mm1 \n\t" - "psubb %%mm1, %%mm2 \n\t" - "paddb %%mm7, %%mm2 \n\t" - "pcmpgtb %%mm6, %%mm2 \n\t" - "paddb %%mm2, %%mm0 \n\t" - "psubusb %%mm3, %%mm4 \n\t" - - "pxor %%mm6, %%mm6 \n\t" - "movq %4, %%mm7 \n\t" // QP,..., QP - "paddusb %%mm7, %%mm7 \n\t" // 2QP ... 2QP - "psubusb %%mm4, %%mm7 \n\t" // Diff >=2QP -> 0 - "pcmpeqb %%mm6, %%mm7 \n\t" // Diff < 2QP -> 0 - "pcmpeqb %%mm6, %%mm7 \n\t" // Diff < 2QP -> 0 - "movq %%mm7, %1 \n\t" - - "movq %5, %%mm7 \n\t" - "punpcklbw %%mm7, %%mm7 \n\t" - "punpcklbw %%mm7, %%mm7 \n\t" - "punpcklbw %%mm7, %%mm7 \n\t" - "psubb %%mm0, %%mm6 \n\t" - "pcmpgtb %%mm7, %%mm6 \n\t" - "movq %%mm6, %0 \n\t" - - : "=m" (eq_mask), "=m" (dc_mask) - : "r" (src), "r" ((long)step), "m" (c->pQPb), "m"(c->ppMode.flatnessThreshold) - : "%"REG_a - ); - - both_masks = dc_mask & eq_mask; - - if(both_masks){ - long offset= -8*step; - int64_t *temp_sums= sums; - - asm volatile( - "movq %2, %%mm0 \n\t" // QP,..., QP - "pxor %%mm4, %%mm4 \n\t" - - "movq (%0), %%mm6 \n\t" - "movq (%0, %1), %%mm5 \n\t" - "movq %%mm5, %%mm1 \n\t" - "movq %%mm6, %%mm2 \n\t" - "psubusb %%mm6, %%mm5 \n\t" - "psubusb %%mm1, %%mm2 \n\t" - "por %%mm5, %%mm2 \n\t" // ABS Diff of lines - "psubusb %%mm2, %%mm0 \n\t" // diff >= QP -> 0 - "pcmpeqb %%mm4, %%mm0 \n\t" // diff >= QP -> FF - - "pxor %%mm6, %%mm1 \n\t" - "pand %%mm0, %%mm1 \n\t" - "pxor %%mm1, %%mm6 \n\t" - // 0:QP 6:First - - "movq (%0, %1, 8), %%mm5 \n\t" - "add %1, %0 \n\t" // %0 points to line 1 not 0 - "movq (%0, %1, 8), %%mm7 \n\t" - "movq %%mm5, %%mm1 \n\t" - "movq %%mm7, %%mm2 \n\t" - "psubusb %%mm7, %%mm5 \n\t" - "psubusb %%mm1, %%mm2 \n\t" - "por %%mm5, %%mm2 \n\t" // ABS Diff of lines - "movq %2, %%mm0 \n\t" // QP,..., QP - "psubusb %%mm2, %%mm0 \n\t" // diff >= QP -> 0 - "pcmpeqb %%mm4, %%mm0 \n\t" // diff >= QP -> FF - - "pxor %%mm7, %%mm1 \n\t" - "pand %%mm0, %%mm1 \n\t" - "pxor %%mm1, %%mm7 \n\t" - - "movq %%mm6, %%mm5 \n\t" - "punpckhbw %%mm4, %%mm6 \n\t" - "punpcklbw %%mm4, %%mm5 \n\t" - // 4:0 5/6:First 7:Last - - "movq %%mm5, %%mm0 \n\t" - "movq %%mm6, %%mm1 \n\t" - "psllw $2, %%mm0 \n\t" - "psllw $2, %%mm1 \n\t" - "paddw "MANGLE(w04)", %%mm0 \n\t" - "paddw "MANGLE(w04)", %%mm1 \n\t" + "psubb %%mm2, %%mm1 \n\t" + "paddb %%mm7, %%mm1 \n\t" + "pcmpgtb %%mm6, %%mm1 \n\t" + "paddb %%mm1, %%mm0 \n\t" + + "movq (%%"REG_a", %3, 4), %%mm1 \n\t" + "psubb %%mm1, %%mm2 \n\t" + "paddb %%mm7, %%mm2 \n\t" + "pcmpgtb %%mm6, %%mm2 \n\t" + "paddb %%mm2, %%mm0 \n\t" + "psubusb %%mm3, %%mm4 \n\t" + + "pxor %%mm6, %%mm6 \n\t" + "movq %4, %%mm7 \n\t" // QP,..., QP + "paddusb %%mm7, %%mm7 \n\t" // 2QP ... 2QP + "psubusb %%mm4, %%mm7 \n\t" // Diff >=2QP -> 0 + "pcmpeqb %%mm6, %%mm7 \n\t" // Diff < 2QP -> 0 + "pcmpeqb %%mm6, %%mm7 \n\t" // Diff < 2QP -> 0 + "movq %%mm7, %1 \n\t" + + "movq %5, %%mm7 \n\t" + "punpcklbw %%mm7, %%mm7 \n\t" + "punpcklbw %%mm7, %%mm7 \n\t" + "punpcklbw %%mm7, %%mm7 \n\t" + "psubb %%mm0, %%mm6 \n\t" + "pcmpgtb %%mm7, %%mm6 \n\t" + "movq %%mm6, %0 \n\t" + + : "=m" (eq_mask), "=m" (dc_mask) + : "r" (src), "r" ((long)step), "m" (c->pQPb), "m"(c->ppMode.flatnessThreshold) + : "%"REG_a + ); + + both_masks = dc_mask & eq_mask; + + if(both_masks){ + long offset= -8*step; + int64_t *temp_sums= sums; + + asm volatile( + "movq %2, %%mm0 \n\t" // QP,..., QP + "pxor %%mm4, %%mm4 \n\t" + + "movq (%0), %%mm6 \n\t" + "movq (%0, %1), %%mm5 \n\t" + "movq %%mm5, %%mm1 \n\t" + "movq %%mm6, %%mm2 \n\t" + "psubusb %%mm6, %%mm5 \n\t" + "psubusb %%mm1, %%mm2 \n\t" + "por %%mm5, %%mm2 \n\t" // ABS Diff of lines + "psubusb %%mm2, %%mm0 \n\t" // diff >= QP -> 0 + "pcmpeqb %%mm4, %%mm0 \n\t" // diff >= QP -> FF + + "pxor %%mm6, %%mm1 \n\t" + "pand %%mm0, %%mm1 \n\t" + "pxor %%mm1, %%mm6 \n\t" + // 0:QP 6:First + + "movq (%0, %1, 8), %%mm5 \n\t" + "add %1, %0 \n\t" // %0 points to line 1 not 0 + "movq (%0, %1, 8), %%mm7 \n\t" + "movq %%mm5, %%mm1 \n\t" + "movq %%mm7, %%mm2 \n\t" + "psubusb %%mm7, %%mm5 \n\t" + "psubusb %%mm1, %%mm2 \n\t" + "por %%mm5, %%mm2 \n\t" // ABS Diff of lines + "movq %2, %%mm0 \n\t" // QP,..., QP + "psubusb %%mm2, %%mm0 \n\t" // diff >= QP -> 0 + "pcmpeqb %%mm4, %%mm0 \n\t" // diff >= QP -> FF + + "pxor %%mm7, %%mm1 \n\t" + "pand %%mm0, %%mm1 \n\t" + "pxor %%mm1, %%mm7 \n\t" + + "movq %%mm6, %%mm5 \n\t" + "punpckhbw %%mm4, %%mm6 \n\t" + "punpcklbw %%mm4, %%mm5 \n\t" + // 4:0 5/6:First 7:Last + + "movq %%mm5, %%mm0 \n\t" + "movq %%mm6, %%mm1 \n\t" + "psllw $2, %%mm0 \n\t" + "psllw $2, %%mm1 \n\t" + "paddw "MANGLE(w04)", %%mm0 \n\t" + "paddw "MANGLE(w04)", %%mm1 \n\t" #define NEXT\ - "movq (%0), %%mm2 \n\t"\ - "movq (%0), %%mm3 \n\t"\ - "add %1, %0 \n\t"\ - "punpcklbw %%mm4, %%mm2 \n\t"\ - "punpckhbw %%mm4, %%mm3 \n\t"\ - "paddw %%mm2, %%mm0 \n\t"\ - "paddw %%mm3, %%mm1 \n\t" + "movq (%0), %%mm2 \n\t"\ + "movq (%0), %%mm3 \n\t"\ + "add %1, %0 \n\t"\ + "punpcklbw %%mm4, %%mm2 \n\t"\ + "punpckhbw %%mm4, %%mm3 \n\t"\ + "paddw %%mm2, %%mm0 \n\t"\ + "paddw %%mm3, %%mm1 \n\t" #define PREV\ - "movq (%0), %%mm2 \n\t"\ - "movq (%0), %%mm3 \n\t"\ - "add %1, %0 \n\t"\ - "punpcklbw %%mm4, %%mm2 \n\t"\ - "punpckhbw %%mm4, %%mm3 \n\t"\ - "psubw %%mm2, %%mm0 \n\t"\ - "psubw %%mm3, %%mm1 \n\t" - - - NEXT //0 - NEXT //1 - NEXT //2 - "movq %%mm0, (%3) \n\t" - "movq %%mm1, 8(%3) \n\t" - - NEXT //3 - "psubw %%mm5, %%mm0 \n\t" - "psubw %%mm6, %%mm1 \n\t" - "movq %%mm0, 16(%3) \n\t" - "movq %%mm1, 24(%3) \n\t" - - NEXT //4 - "psubw %%mm5, %%mm0 \n\t" - "psubw %%mm6, %%mm1 \n\t" - "movq %%mm0, 32(%3) \n\t" - "movq %%mm1, 40(%3) \n\t" - - NEXT //5 - "psubw %%mm5, %%mm0 \n\t" - "psubw %%mm6, %%mm1 \n\t" - "movq %%mm0, 48(%3) \n\t" - "movq %%mm1, 56(%3) \n\t" - - NEXT //6 - "psubw %%mm5, %%mm0 \n\t" - "psubw %%mm6, %%mm1 \n\t" - "movq %%mm0, 64(%3) \n\t" - "movq %%mm1, 72(%3) \n\t" - - "movq %%mm7, %%mm6 \n\t" - "punpckhbw %%mm4, %%mm7 \n\t" - "punpcklbw %%mm4, %%mm6 \n\t" - - NEXT //7 - "mov %4, %0 \n\t" - "add %1, %0 \n\t" - PREV //0 - "movq %%mm0, 80(%3) \n\t" - "movq %%mm1, 88(%3) \n\t" - - PREV //1 - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" - "movq %%mm0, 96(%3) \n\t" - "movq %%mm1, 104(%3) \n\t" - - PREV //2 - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" - "movq %%mm0, 112(%3) \n\t" - "movq %%mm1, 120(%3) \n\t" - - PREV //3 - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" - "movq %%mm0, 128(%3) \n\t" - "movq %%mm1, 136(%3) \n\t" - - PREV //4 - "paddw %%mm6, %%mm0 \n\t" - "paddw %%mm7, %%mm1 \n\t" - "movq %%mm0, 144(%3) \n\t" - "movq %%mm1, 152(%3) \n\t" - - "mov %4, %0 \n\t" //FIXME - - : "+&r"(src) - : "r" ((long)step), "m" (c->pQPb), "r"(sums), "g"(src) - ); - - src+= step; // src points to begin of the 8x8 Block - - asm volatile( - "movq %4, %%mm6 \n\t" - "pcmpeqb %%mm5, %%mm5 \n\t" - "pxor %%mm6, %%mm5 \n\t" - "pxor %%mm7, %%mm7 \n\t" - - "1: \n\t" - "movq (%1), %%mm0 \n\t" - "movq 8(%1), %%mm1 \n\t" - "paddw 32(%1), %%mm0 \n\t" - "paddw 40(%1), %%mm1 \n\t" - "movq (%0, %3), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "movq %%mm2, %%mm4 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" - "punpckhbw %%mm7, %%mm3 \n\t" - "paddw %%mm2, %%mm0 \n\t" - "paddw %%mm3, %%mm1 \n\t" - "paddw %%mm2, %%mm0 \n\t" - "paddw %%mm3, %%mm1 \n\t" - "psrlw $4, %%mm0 \n\t" - "psrlw $4, %%mm1 \n\t" - "packuswb %%mm1, %%mm0 \n\t" - "pand %%mm6, %%mm0 \n\t" - "pand %%mm5, %%mm4 \n\t" - "por %%mm4, %%mm0 \n\t" - "movq %%mm0, (%0, %3) \n\t" - "add $16, %1 \n\t" - "add %2, %0 \n\t" - " js 1b \n\t" - - : "+r"(offset), "+r"(temp_sums) - : "r" ((long)step), "r"(src - offset), "m"(both_masks) - ); - }else - src+= step; // src points to begin of the 8x8 Block - - if(eq_mask != -1LL){ - uint8_t *temp_src= src; - asm volatile( - "pxor %%mm7, %%mm7 \n\t" - "lea -40(%%"REG_SP"), %%"REG_c" \n\t" // make space for 4 8-byte vars - "and "ALIGN_MASK", %%"REG_c" \n\t" // align -// 0 1 2 3 4 5 6 7 8 9 -// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %1+8%1 ecx+4%1 - - "movq (%0), %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" // low part of line 0 - "punpckhbw %%mm7, %%mm1 \n\t" // high part of line 0 - - "movq (%0, %1), %%mm2 \n\t" - "lea (%0, %1, 2), %%"REG_a" \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // low part of line 1 - "punpckhbw %%mm7, %%mm3 \n\t" // high part of line 1 - - "movq (%%"REG_a"), %%mm4 \n\t" - "movq %%mm4, %%mm5 \n\t" - "punpcklbw %%mm7, %%mm4 \n\t" // low part of line 2 - "punpckhbw %%mm7, %%mm5 \n\t" // high part of line 2 - - "paddw %%mm0, %%mm0 \n\t" // 2L0 - "paddw %%mm1, %%mm1 \n\t" // 2H0 - "psubw %%mm4, %%mm2 \n\t" // L1 - L2 - "psubw %%mm5, %%mm3 \n\t" // H1 - H2 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - L1 + L2 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - H1 + H2 - - "psllw $2, %%mm2 \n\t" // 4L1 - 4L2 - "psllw $2, %%mm3 \n\t" // 4H1 - 4H2 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - - "movq (%%"REG_a", %1), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L3 - "punpckhbw %%mm7, %%mm3 \n\t" // H3 - - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - L3 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - H3 - "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 - "movq %%mm0, (%%"REG_c") \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "movq %%mm1, 8(%%"REG_c") \n\t" // 2H0 - 5H1 + 5H2 - 2H3 - - "movq (%%"REG_a", %1, 2), %%mm0 \n\t" - "movq %%mm0, %%mm1 \n\t" - "punpcklbw %%mm7, %%mm0 \n\t" // L4 - "punpckhbw %%mm7, %%mm1 \n\t" // H4 - - "psubw %%mm0, %%mm2 \n\t" // L3 - L4 - "psubw %%mm1, %%mm3 \n\t" // H3 - H4 - "movq %%mm2, 16(%%"REG_c") \n\t" // L3 - L4 - "movq %%mm3, 24(%%"REG_c") \n\t" // H3 - H4 - "paddw %%mm4, %%mm4 \n\t" // 2L2 - "paddw %%mm5, %%mm5 \n\t" // 2H2 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - L3 + L4 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - H3 + H4 - - "lea (%%"REG_a", %1), %0 \n\t" - "psllw $2, %%mm2 \n\t" // 4L3 - 4L4 - "psllw $2, %%mm3 \n\t" // 4H3 - 4H4 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 + "movq (%0), %%mm2 \n\t"\ + "movq (%0), %%mm3 \n\t"\ + "add %1, %0 \n\t"\ + "punpcklbw %%mm4, %%mm2 \n\t"\ + "punpckhbw %%mm4, %%mm3 \n\t"\ + "psubw %%mm2, %%mm0 \n\t"\ + "psubw %%mm3, %%mm1 \n\t" + + + NEXT //0 + NEXT //1 + NEXT //2 + "movq %%mm0, (%3) \n\t" + "movq %%mm1, 8(%3) \n\t" + + NEXT //3 + "psubw %%mm5, %%mm0 \n\t" + "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, 16(%3) \n\t" + "movq %%mm1, 24(%3) \n\t" + + NEXT //4 + "psubw %%mm5, %%mm0 \n\t" + "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, 32(%3) \n\t" + "movq %%mm1, 40(%3) \n\t" + + NEXT //5 + "psubw %%mm5, %%mm0 \n\t" + "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, 48(%3) \n\t" + "movq %%mm1, 56(%3) \n\t" + + NEXT //6 + "psubw %%mm5, %%mm0 \n\t" + "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, 64(%3) \n\t" + "movq %%mm1, 72(%3) \n\t" + + "movq %%mm7, %%mm6 \n\t" + "punpckhbw %%mm4, %%mm7 \n\t" + "punpcklbw %%mm4, %%mm6 \n\t" + + NEXT //7 + "mov %4, %0 \n\t" + "add %1, %0 \n\t" + PREV //0 + "movq %%mm0, 80(%3) \n\t" + "movq %%mm1, 88(%3) \n\t" + + PREV //1 + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" + "movq %%mm0, 96(%3) \n\t" + "movq %%mm1, 104(%3) \n\t" + + PREV //2 + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" + "movq %%mm0, 112(%3) \n\t" + "movq %%mm1, 120(%3) \n\t" + + PREV //3 + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" + "movq %%mm0, 128(%3) \n\t" + "movq %%mm1, 136(%3) \n\t" + + PREV //4 + "paddw %%mm6, %%mm0 \n\t" + "paddw %%mm7, %%mm1 \n\t" + "movq %%mm0, 144(%3) \n\t" + "movq %%mm1, 152(%3) \n\t" + + "mov %4, %0 \n\t" //FIXME + + : "+&r"(src) + : "r" ((long)step), "m" (c->pQPb), "r"(sums), "g"(src) + ); + + src+= step; // src points to begin of the 8x8 Block + + asm volatile( + "movq %4, %%mm6 \n\t" + "pcmpeqb %%mm5, %%mm5 \n\t" + "pxor %%mm6, %%mm5 \n\t" + "pxor %%mm7, %%mm7 \n\t" + + "1: \n\t" + "movq (%1), %%mm0 \n\t" + "movq 8(%1), %%mm1 \n\t" + "paddw 32(%1), %%mm0 \n\t" + "paddw 40(%1), %%mm1 \n\t" + "movq (%0, %3), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "movq %%mm2, %%mm4 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "paddw %%mm2, %%mm0 \n\t" + "paddw %%mm3, %%mm1 \n\t" + "paddw %%mm2, %%mm0 \n\t" + "paddw %%mm3, %%mm1 \n\t" + "psrlw $4, %%mm0 \n\t" + "psrlw $4, %%mm1 \n\t" + "packuswb %%mm1, %%mm0 \n\t" + "pand %%mm6, %%mm0 \n\t" + "pand %%mm5, %%mm4 \n\t" + "por %%mm4, %%mm0 \n\t" + "movq %%mm0, (%0, %3) \n\t" + "add $16, %1 \n\t" + "add %2, %0 \n\t" + " js 1b \n\t" + + : "+r"(offset), "+r"(temp_sums) + : "r" ((long)step), "r"(src - offset), "m"(both_masks) + ); + }else + src+= step; // src points to begin of the 8x8 Block + + if(eq_mask != -1LL){ + uint8_t *temp_src= src; + asm volatile( + "pxor %%mm7, %%mm7 \n\t" + "lea -40(%%"REG_SP"), %%"REG_c" \n\t" // make space for 4 8-byte vars + "and "ALIGN_MASK", %%"REG_c" \n\t" // align +// 0 1 2 3 4 5 6 7 8 9 +// %0 eax eax+%1 eax+2%1 %0+4%1 ecx ecx+%1 ecx+2%1 %1+8%1 ecx+4%1 + + "movq (%0), %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" // low part of line 0 + "punpckhbw %%mm7, %%mm1 \n\t" // high part of line 0 + + "movq (%0, %1), %%mm2 \n\t" + "lea (%0, %1, 2), %%"REG_a" \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // low part of line 1 + "punpckhbw %%mm7, %%mm3 \n\t" // high part of line 1 + + "movq (%%"REG_a"), %%mm4 \n\t" + "movq %%mm4, %%mm5 \n\t" + "punpcklbw %%mm7, %%mm4 \n\t" // low part of line 2 + "punpckhbw %%mm7, %%mm5 \n\t" // high part of line 2 + + "paddw %%mm0, %%mm0 \n\t" // 2L0 + "paddw %%mm1, %%mm1 \n\t" // 2H0 + "psubw %%mm4, %%mm2 \n\t" // L1 - L2 + "psubw %%mm5, %%mm3 \n\t" // H1 - H2 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - L1 + L2 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - H1 + H2 + + "psllw $2, %%mm2 \n\t" // 4L1 - 4L2 + "psllw $2, %%mm3 \n\t" // 4H1 - 4H2 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 + + "movq (%%"REG_a", %1), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L3 + "punpckhbw %%mm7, %%mm3 \n\t" // H3 + + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - L3 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - H3 + "psubw %%mm2, %%mm0 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "psubw %%mm3, %%mm1 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + "movq %%mm0, (%%"REG_c") \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "movq %%mm1, 8(%%"REG_c") \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + + "movq (%%"REG_a", %1, 2), %%mm0 \n\t" + "movq %%mm0, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" // L4 + "punpckhbw %%mm7, %%mm1 \n\t" // H4 + + "psubw %%mm0, %%mm2 \n\t" // L3 - L4 + "psubw %%mm1, %%mm3 \n\t" // H3 - H4 + "movq %%mm2, 16(%%"REG_c") \n\t" // L3 - L4 + "movq %%mm3, 24(%%"REG_c") \n\t" // H3 - H4 + "paddw %%mm4, %%mm4 \n\t" // 2L2 + "paddw %%mm5, %%mm5 \n\t" // 2H2 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - L3 + L4 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - H3 + H4 + + "lea (%%"REG_a", %1), %0 \n\t" + "psllw $2, %%mm2 \n\t" // 4L3 - 4L4 + "psllw $2, %%mm3 \n\t" // 4H3 - 4H4 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 //50 opcodes so far - "movq (%0, %1, 2), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L5 - "punpckhbw %%mm7, %%mm3 \n\t" // H5 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - L5 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - H5 - "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - 2L5 - "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - 2H5 - - "movq (%%"REG_a", %1, 4), %%mm6 \n\t" - "punpcklbw %%mm7, %%mm6 \n\t" // L6 - "psubw %%mm6, %%mm2 \n\t" // L5 - L6 - "movq (%%"REG_a", %1, 4), %%mm6 \n\t" - "punpckhbw %%mm7, %%mm6 \n\t" // H6 - "psubw %%mm6, %%mm3 \n\t" // H5 - H6 - - "paddw %%mm0, %%mm0 \n\t" // 2L4 - "paddw %%mm1, %%mm1 \n\t" // 2H4 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - L5 + L6 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - H5 + H6 - - "psllw $2, %%mm2 \n\t" // 4L5 - 4L6 - "psllw $2, %%mm3 \n\t" // 4H5 - 4H6 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - - "movq (%0, %1, 4), %%mm2 \n\t" - "movq %%mm2, %%mm3 \n\t" - "punpcklbw %%mm7, %%mm2 \n\t" // L7 - "punpckhbw %%mm7, %%mm3 \n\t" // H7 - - "paddw %%mm2, %%mm2 \n\t" // 2L7 - "paddw %%mm3, %%mm3 \n\t" // 2H7 - "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - 2L7 - "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - 2H7 - - "movq (%%"REG_c"), %%mm2 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 - "movq 8(%%"REG_c"), %%mm3 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 + "movq (%0, %1, 2), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L5 + "punpckhbw %%mm7, %%mm3 \n\t" // H5 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - L5 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - H5 + "psubw %%mm2, %%mm4 \n\t" // 2L2 - 5L3 + 5L4 - 2L5 + "psubw %%mm3, %%mm5 \n\t" // 2H2 - 5H3 + 5H4 - 2H5 + + "movq (%%"REG_a", %1, 4), %%mm6 \n\t" + "punpcklbw %%mm7, %%mm6 \n\t" // L6 + "psubw %%mm6, %%mm2 \n\t" // L5 - L6 + "movq (%%"REG_a", %1, 4), %%mm6 \n\t" + "punpckhbw %%mm7, %%mm6 \n\t" // H6 + "psubw %%mm6, %%mm3 \n\t" // H5 - H6 + + "paddw %%mm0, %%mm0 \n\t" // 2L4 + "paddw %%mm1, %%mm1 \n\t" // 2H4 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - L5 + L6 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - H5 + H6 + + "psllw $2, %%mm2 \n\t" // 4L5 - 4L6 + "psllw $2, %%mm3 \n\t" // 4H5 - 4H6 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 + + "movq (%0, %1, 4), %%mm2 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" // L7 + "punpckhbw %%mm7, %%mm3 \n\t" // H7 + + "paddw %%mm2, %%mm2 \n\t" // 2L7 + "paddw %%mm3, %%mm3 \n\t" // 2H7 + "psubw %%mm2, %%mm0 \n\t" // 2L4 - 5L5 + 5L6 - 2L7 + "psubw %%mm3, %%mm1 \n\t" // 2H4 - 5H5 + 5H6 - 2H7 + + "movq (%%"REG_c"), %%mm2 \n\t" // 2L0 - 5L1 + 5L2 - 2L3 + "movq 8(%%"REG_c"), %%mm3 \n\t" // 2H0 - 5H1 + 5H2 - 2H3 #ifdef HAVE_MMX2 - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm0, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm1, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm2, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| - "movq %%mm7, %%mm6 \n\t" // 0 - "psubw %%mm3, %%mm6 \n\t" - "pmaxsw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm0, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm1, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm2, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| + "movq %%mm7, %%mm6 \n\t" // 0 + "psubw %%mm3, %%mm6 \n\t" + "pmaxsw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| #else - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm0, %%mm6 \n\t" - "pxor %%mm6, %%mm0 \n\t" - "psubw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm1, %%mm6 \n\t" - "pxor %%mm6, %%mm1 \n\t" - "psubw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm2, %%mm6 \n\t" - "pxor %%mm6, %%mm2 \n\t" - "psubw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm3, %%mm6 \n\t" - "pxor %%mm6, %%mm3 \n\t" - "psubw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm0, %%mm6 \n\t" + "pxor %%mm6, %%mm0 \n\t" + "psubw %%mm6, %%mm0 \n\t" // |2L4 - 5L5 + 5L6 - 2L7| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm1, %%mm6 \n\t" + "pxor %%mm6, %%mm1 \n\t" + "psubw %%mm6, %%mm1 \n\t" // |2H4 - 5H5 + 5H6 - 2H7| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm2, %%mm6 \n\t" + "pxor %%mm6, %%mm2 \n\t" + "psubw %%mm6, %%mm2 \n\t" // |2L0 - 5L1 + 5L2 - 2L3| + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm3, %%mm6 \n\t" + "pxor %%mm6, %%mm3 \n\t" + "psubw %%mm6, %%mm3 \n\t" // |2H0 - 5H1 + 5H2 - 2H3| #endif #ifdef HAVE_MMX2 - "pminsw %%mm2, %%mm0 \n\t" - "pminsw %%mm3, %%mm1 \n\t" + "pminsw %%mm2, %%mm0 \n\t" + "pminsw %%mm3, %%mm1 \n\t" #else - "movq %%mm0, %%mm6 \n\t" - "psubusw %%mm2, %%mm6 \n\t" - "psubw %%mm6, %%mm0 \n\t" - "movq %%mm1, %%mm6 \n\t" - "psubusw %%mm3, %%mm6 \n\t" - "psubw %%mm6, %%mm1 \n\t" + "movq %%mm0, %%mm6 \n\t" + "psubusw %%mm2, %%mm6 \n\t" + "psubw %%mm6, %%mm0 \n\t" + "movq %%mm1, %%mm6 \n\t" + "psubusw %%mm3, %%mm6 \n\t" + "psubw %%mm6, %%mm1 \n\t" #endif - "movd %2, %%mm2 \n\t" // QP - "punpcklbw %%mm7, %%mm2 \n\t" - - "movq %%mm7, %%mm6 \n\t" // 0 - "pcmpgtw %%mm4, %%mm6 \n\t" // sign(2L2 - 5L3 + 5L4 - 2L5) - "pxor %%mm6, %%mm4 \n\t" - "psubw %%mm6, %%mm4 \n\t" // |2L2 - 5L3 + 5L4 - 2L5| - "pcmpgtw %%mm5, %%mm7 \n\t" // sign(2H2 - 5H3 + 5H4 - 2H5) - "pxor %%mm7, %%mm5 \n\t" - "psubw %%mm7, %%mm5 \n\t" // |2H2 - 5H3 + 5H4 - 2H5| + "movd %2, %%mm2 \n\t" // QP + "punpcklbw %%mm7, %%mm2 \n\t" + + "movq %%mm7, %%mm6 \n\t" // 0 + "pcmpgtw %%mm4, %%mm6 \n\t" // sign(2L2 - 5L3 + 5L4 - 2L5) + "pxor %%mm6, %%mm4 \n\t" + "psubw %%mm6, %%mm4 \n\t" // |2L2 - 5L3 + 5L4 - 2L5| + "pcmpgtw %%mm5, %%mm7 \n\t" // sign(2H2 - 5H3 + 5H4 - 2H5) + "pxor %%mm7, %%mm5 \n\t" + "psubw %%mm7, %%mm5 \n\t" // |2H2 - 5H3 + 5H4 - 2H5| // 100 opcodes - "psllw $3, %%mm2 \n\t" // 8QP - "movq %%mm2, %%mm3 \n\t" // 8QP - "pcmpgtw %%mm4, %%mm2 \n\t" - "pcmpgtw %%mm5, %%mm3 \n\t" - "pand %%mm2, %%mm4 \n\t" - "pand %%mm3, %%mm5 \n\t" - - - "psubusw %%mm0, %%mm4 \n\t" // hd - "psubusw %%mm1, %%mm5 \n\t" // ld - - - "movq "MANGLE(w05)", %%mm2 \n\t" // 5 - "pmullw %%mm2, %%mm4 \n\t" - "pmullw %%mm2, %%mm5 \n\t" - "movq "MANGLE(w20)", %%mm2 \n\t" // 32 - "paddw %%mm2, %%mm4 \n\t" - "paddw %%mm2, %%mm5 \n\t" - "psrlw $6, %%mm4 \n\t" - "psrlw $6, %%mm5 \n\t" - - "movq 16(%%"REG_c"), %%mm0 \n\t" // L3 - L4 - "movq 24(%%"REG_c"), %%mm1 \n\t" // H3 - H4 - - "pxor %%mm2, %%mm2 \n\t" - "pxor %%mm3, %%mm3 \n\t" - - "pcmpgtw %%mm0, %%mm2 \n\t" // sign (L3-L4) - "pcmpgtw %%mm1, %%mm3 \n\t" // sign (H3-H4) - "pxor %%mm2, %%mm0 \n\t" - "pxor %%mm3, %%mm1 \n\t" - "psubw %%mm2, %%mm0 \n\t" // |L3-L4| - "psubw %%mm3, %%mm1 \n\t" // |H3-H4| - "psrlw $1, %%mm0 \n\t" // |L3 - L4|/2 - "psrlw $1, %%mm1 \n\t" // |H3 - H4|/2 - - "pxor %%mm6, %%mm2 \n\t" - "pxor %%mm7, %%mm3 \n\t" - "pand %%mm2, %%mm4 \n\t" - "pand %%mm3, %%mm5 \n\t" + "psllw $3, %%mm2 \n\t" // 8QP + "movq %%mm2, %%mm3 \n\t" // 8QP + "pcmpgtw %%mm4, %%mm2 \n\t" + "pcmpgtw %%mm5, %%mm3 \n\t" + "pand %%mm2, %%mm4 \n\t" + "pand %%mm3, %%mm5 \n\t" + + + "psubusw %%mm0, %%mm4 \n\t" // hd + "psubusw %%mm1, %%mm5 \n\t" // ld + + + "movq "MANGLE(w05)", %%mm2 \n\t" // 5 + "pmullw %%mm2, %%mm4 \n\t" + "pmullw %%mm2, %%mm5 \n\t" + "movq "MANGLE(w20)", %%mm2 \n\t" // 32 + "paddw %%mm2, %%mm4 \n\t" + "paddw %%mm2, %%mm5 \n\t" + "psrlw $6, %%mm4 \n\t" + "psrlw $6, %%mm5 \n\t" + + "movq 16(%%"REG_c"), %%mm0 \n\t" // L3 - L4 + "movq 24(%%"REG_c"), %%mm1 \n\t" // H3 - H4 + + "pxor %%mm2, %%mm2 \n\t" + "pxor %%mm3, %%mm3 \n\t" + + "pcmpgtw %%mm0, %%mm2 \n\t" // sign (L3-L4) + "pcmpgtw %%mm1, %%mm3 \n\t" // sign (H3-H4) + "pxor %%mm2, %%mm0 \n\t" + "pxor %%mm3, %%mm1 \n\t" + "psubw %%mm2, %%mm0 \n\t" // |L3-L4| + "psubw %%mm3, %%mm1 \n\t" // |H3-H4| + "psrlw $1, %%mm0 \n\t" // |L3 - L4|/2 + "psrlw $1, %%mm1 \n\t" // |H3 - H4|/2 + + "pxor %%mm6, %%mm2 \n\t" + "pxor %%mm7, %%mm3 \n\t" + "pand %%mm2, %%mm4 \n\t" + "pand %%mm3, %%mm5 \n\t" #ifdef HAVE_MMX2 - "pminsw %%mm0, %%mm4 \n\t" - "pminsw %%mm1, %%mm5 \n\t" + "pminsw %%mm0, %%mm4 \n\t" + "pminsw %%mm1, %%mm5 \n\t" #else - "movq %%mm4, %%mm2 \n\t" - "psubusw %%mm0, %%mm2 \n\t" - "psubw %%mm2, %%mm4 \n\t" - "movq %%mm5, %%mm2 \n\t" - "psubusw %%mm1, %%mm2 \n\t" - "psubw %%mm2, %%mm5 \n\t" + "movq %%mm4, %%mm2 \n\t" + "psubusw %%mm0, %%mm2 \n\t" + "psubw %%mm2, %%mm4 \n\t" + "movq %%mm5, %%mm2 \n\t" + "psubusw %%mm1, %%mm2 \n\t" + "psubw %%mm2, %%mm5 \n\t" #endif - "pxor %%mm6, %%mm4 \n\t" - "pxor %%mm7, %%mm5 \n\t" - "psubw %%mm6, %%mm4 \n\t" - "psubw %%mm7, %%mm5 \n\t" - "packsswb %%mm5, %%mm4 \n\t" - "movq %3, %%mm1 \n\t" - "pandn %%mm4, %%mm1 \n\t" - "movq (%0), %%mm0 \n\t" - "paddb %%mm1, %%mm0 \n\t" - "movq %%mm0, (%0) \n\t" - "movq (%0, %1), %%mm0 \n\t" - "psubb %%mm1, %%mm0 \n\t" - "movq %%mm0, (%0, %1) \n\t" - - : "+r" (temp_src) - : "r" ((long)step), "m" (c->pQPb), "m"(eq_mask) - : "%"REG_a, "%"REG_c - ); - } + "pxor %%mm6, %%mm4 \n\t" + "pxor %%mm7, %%mm5 \n\t" + "psubw %%mm6, %%mm4 \n\t" + "psubw %%mm7, %%mm5 \n\t" + "packsswb %%mm5, %%mm4 \n\t" + "movq %3, %%mm1 \n\t" + "pandn %%mm4, %%mm1 \n\t" + "movq (%0), %%mm0 \n\t" + "paddb %%mm1, %%mm0 \n\t" + "movq %%mm0, (%0) \n\t" + "movq (%0, %1), %%mm0 \n\t" + "psubb %%mm1, %%mm0 \n\t" + "movq %%mm0, (%0, %1) \n\t" + + : "+r" (temp_src) + : "r" ((long)step), "m" (c->pQPb), "m"(eq_mask) + : "%"REG_a, "%"REG_c + ); + } /*if(step==16){ STOP_TIMER("step16") }else{ @@ -3180,7 +3180,7 @@ #endif //HAVE_MMX static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height, - QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c); + QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c); /** * Copies a block from src to dst and fixes the blacklevel @@ -3189,69 +3189,69 @@ #undef SCALED_CPY static inline void RENAME(blockCopy)(uint8_t dst[], int dstStride, uint8_t src[], int srcStride, - int levelFix, int64_t *packedOffsetAndScale) + int levelFix, int64_t *packedOffsetAndScale) { #ifndef HAVE_MMX - int i; + int i; #endif - if(levelFix) - { + if(levelFix) + { #ifdef HAVE_MMX - asm volatile( - "movq (%%"REG_a"), %%mm2 \n\t" // packedYOffset - "movq 8(%%"REG_a"), %%mm3 \n\t" // packedYScale - "lea (%2,%4), %%"REG_a" \n\t" - "lea (%3,%5), %%"REG_d" \n\t" - "pxor %%mm4, %%mm4 \n\t" + asm volatile( + "movq (%%"REG_a"), %%mm2 \n\t" // packedYOffset + "movq 8(%%"REG_a"), %%mm3 \n\t" // packedYScale + "lea (%2,%4), %%"REG_a" \n\t" + "lea (%3,%5), %%"REG_d" \n\t" + "pxor %%mm4, %%mm4 \n\t" #ifdef HAVE_MMX2 -#define REAL_SCALED_CPY(src1, src2, dst1, dst2) \ - "movq " #src1 ", %%mm0 \n\t"\ - "movq " #src1 ", %%mm5 \n\t"\ - "movq " #src2 ", %%mm1 \n\t"\ - "movq " #src2 ", %%mm6 \n\t"\ - "punpcklbw %%mm0, %%mm0 \n\t"\ - "punpckhbw %%mm5, %%mm5 \n\t"\ - "punpcklbw %%mm1, %%mm1 \n\t"\ - "punpckhbw %%mm6, %%mm6 \n\t"\ - "pmulhuw %%mm3, %%mm0 \n\t"\ - "pmulhuw %%mm3, %%mm5 \n\t"\ - "pmulhuw %%mm3, %%mm1 \n\t"\ - "pmulhuw %%mm3, %%mm6 \n\t"\ - "psubw %%mm2, %%mm0 \n\t"\ - "psubw %%mm2, %%mm5 \n\t"\ - "psubw %%mm2, %%mm1 \n\t"\ - "psubw %%mm2, %%mm6 \n\t"\ - "packuswb %%mm5, %%mm0 \n\t"\ - "packuswb %%mm6, %%mm1 \n\t"\ - "movq %%mm0, " #dst1 " \n\t"\ - "movq %%mm1, " #dst2 " \n\t"\ +#define REAL_SCALED_CPY(src1, src2, dst1, dst2) \ + "movq " #src1 ", %%mm0 \n\t"\ + "movq " #src1 ", %%mm5 \n\t"\ + "movq " #src2 ", %%mm1 \n\t"\ + "movq " #src2 ", %%mm6 \n\t"\ + "punpcklbw %%mm0, %%mm0 \n\t"\ + "punpckhbw %%mm5, %%mm5 \n\t"\ + "punpcklbw %%mm1, %%mm1 \n\t"\ + "punpckhbw %%mm6, %%mm6 \n\t"\ + "pmulhuw %%mm3, %%mm0 \n\t"\ + "pmulhuw %%mm3, %%mm5 \n\t"\ + "pmulhuw %%mm3, %%mm1 \n\t"\ + "pmulhuw %%mm3, %%mm6 \n\t"\ + "psubw %%mm2, %%mm0 \n\t"\ + "psubw %%mm2, %%mm5 \n\t"\ + "psubw %%mm2, %%mm1 \n\t"\ + "psubw %%mm2, %%mm6 \n\t"\ + "packuswb %%mm5, %%mm0 \n\t"\ + "packuswb %%mm6, %%mm1 \n\t"\ + "movq %%mm0, " #dst1 " \n\t"\ + "movq %%mm1, " #dst2 " \n\t"\ #else //HAVE_MMX2 -#define REAL_SCALED_CPY(src1, src2, dst1, dst2) \ - "movq " #src1 ", %%mm0 \n\t"\ - "movq " #src1 ", %%mm5 \n\t"\ - "punpcklbw %%mm4, %%mm0 \n\t"\ - "punpckhbw %%mm4, %%mm5 \n\t"\ - "psubw %%mm2, %%mm0 \n\t"\ - "psubw %%mm2, %%mm5 \n\t"\ - "movq " #src2 ", %%mm1 \n\t"\ - "psllw $6, %%mm0 \n\t"\ - "psllw $6, %%mm5 \n\t"\ - "pmulhw %%mm3, %%mm0 \n\t"\ - "movq " #src2 ", %%mm6 \n\t"\ - "pmulhw %%mm3, %%mm5 \n\t"\ - "punpcklbw %%mm4, %%mm1 \n\t"\ - "punpckhbw %%mm4, %%mm6 \n\t"\ - "psubw %%mm2, %%mm1 \n\t"\ - "psubw %%mm2, %%mm6 \n\t"\ - "psllw $6, %%mm1 \n\t"\ - "psllw $6, %%mm6 \n\t"\ - "pmulhw %%mm3, %%mm1 \n\t"\ - "pmulhw %%mm3, %%mm6 \n\t"\ - "packuswb %%mm5, %%mm0 \n\t"\ - "packuswb %%mm6, %%mm1 \n\t"\ - "movq %%mm0, " #dst1 " \n\t"\ - "movq %%mm1, " #dst2 " \n\t"\ +#define REAL_SCALED_CPY(src1, src2, dst1, dst2) \ + "movq " #src1 ", %%mm0 \n\t"\ + "movq " #src1 ", %%mm5 \n\t"\ + "punpcklbw %%mm4, %%mm0 \n\t"\ + "punpckhbw %%mm4, %%mm5 \n\t"\ + "psubw %%mm2, %%mm0 \n\t"\ + "psubw %%mm2, %%mm5 \n\t"\ + "movq " #src2 ", %%mm1 \n\t"\ + "psllw $6, %%mm0 \n\t"\ + "psllw $6, %%mm5 \n\t"\ + "pmulhw %%mm3, %%mm0 \n\t"\ + "movq " #src2 ", %%mm6 \n\t"\ + "pmulhw %%mm3, %%mm5 \n\t"\ + "punpcklbw %%mm4, %%mm1 \n\t"\ + "punpckhbw %%mm4, %%mm6 \n\t"\ + "psubw %%mm2, %%mm1 \n\t"\ + "psubw %%mm2, %%mm6 \n\t"\ + "psllw $6, %%mm1 \n\t"\ + "psllw $6, %%mm6 \n\t"\ + "pmulhw %%mm3, %%mm1 \n\t"\ + "pmulhw %%mm3, %%mm6 \n\t"\ + "packuswb %%mm5, %%mm0 \n\t"\ + "packuswb %%mm6, %%mm1 \n\t"\ + "movq %%mm0, " #dst1 " \n\t"\ + "movq %%mm1, " #dst2 " \n\t"\ #endif //HAVE_MMX2 #define SCALED_CPY(src1, src2, dst1, dst2)\ @@ -3260,60 +3260,60 @@ SCALED_CPY((%2) , (%2, %4) , (%3) , (%3, %5)) SCALED_CPY((%2, %4, 2), (%%REGa, %4, 2), (%3, %5, 2), (%%REGd, %5, 2)) SCALED_CPY((%2, %4, 4), (%%REGa, %4, 4), (%3, %5, 4), (%%REGd, %5, 4)) - "lea (%%"REG_a",%4,4), %%"REG_a" \n\t" - "lea (%%"REG_d",%5,4), %%"REG_d" \n\t" + "lea (%%"REG_a",%4,4), %%"REG_a" \n\t" + "lea (%%"REG_d",%5,4), %%"REG_d" \n\t" SCALED_CPY((%%REGa, %4), (%%REGa, %4, 2), (%%REGd, %5), (%%REGd, %5, 2)) - : "=&a" (packedOffsetAndScale) - : "0" (packedOffsetAndScale), - "r"(src), - "r"(dst), - "r" ((long)srcStride), - "r" ((long)dstStride) - : "%"REG_d - ); + : "=&a" (packedOffsetAndScale) + : "0" (packedOffsetAndScale), + "r"(src), + "r"(dst), + "r" ((long)srcStride), + "r" ((long)dstStride) + : "%"REG_d + ); #else //HAVE_MMX - for(i=0; i<8; i++) - memcpy( &(dst[dstStride*i]), - &(src[srcStride*i]), BLOCK_SIZE); + for(i=0; i<8; i++) + memcpy( &(dst[dstStride*i]), + &(src[srcStride*i]), BLOCK_SIZE); #endif //HAVE_MMX - } - else - { + } + else + { #ifdef HAVE_MMX - asm volatile( - "lea (%0,%2), %%"REG_a" \n\t" - "lea (%1,%3), %%"REG_d" \n\t" - -#define REAL_SIMPLE_CPY(src1, src2, dst1, dst2) \ - "movq " #src1 ", %%mm0 \n\t"\ - "movq " #src2 ", %%mm1 \n\t"\ - "movq %%mm0, " #dst1 " \n\t"\ - "movq %%mm1, " #dst2 " \n\t"\ + asm volatile( + "lea (%0,%2), %%"REG_a" \n\t" + "lea (%1,%3), %%"REG_d" \n\t" + +#define REAL_SIMPLE_CPY(src1, src2, dst1, dst2) \ + "movq " #src1 ", %%mm0 \n\t"\ + "movq " #src2 ", %%mm1 \n\t"\ + "movq %%mm0, " #dst1 " \n\t"\ + "movq %%mm1, " #dst2 " \n\t"\ #define SIMPLE_CPY(src1, src2, dst1, dst2)\ REAL_SIMPLE_CPY(src1, src2, dst1, dst2) -SIMPLE_CPY((%0) , (%0, %2) , (%1) , (%1, %3)) +SIMPLE_CPY((%0) , (%0, %2) , (%1) , (%1, %3)) SIMPLE_CPY((%0, %2, 2), (%%REGa, %2, 2), (%1, %3, 2), (%%REGd, %3, 2)) SIMPLE_CPY((%0, %2, 4), (%%REGa, %2, 4), (%1, %3, 4), (%%REGd, %3, 4)) - "lea (%%"REG_a",%2,4), %%"REG_a" \n\t" - "lea (%%"REG_d",%3,4), %%"REG_d" \n\t" + "lea (%%"REG_a",%2,4), %%"REG_a" \n\t" + "lea (%%"REG_d",%3,4), %%"REG_d" \n\t" SIMPLE_CPY((%%REGa, %2), (%%REGa, %2, 2), (%%REGd, %3), (%%REGd, %3, 2)) - : : "r" (src), - "r" (dst), - "r" ((long)srcStride), - "r" ((long)dstStride) - : "%"REG_a, "%"REG_d - ); + : : "r" (src), + "r" (dst), + "r" ((long)srcStride), + "r" ((long)dstStride) + : "%"REG_a, "%"REG_d + ); #else //HAVE_MMX - for(i=0; i<8; i++) - memcpy( &(dst[dstStride*i]), - &(src[srcStride*i]), BLOCK_SIZE); + for(i=0; i<8; i++) + memcpy( &(dst[dstStride*i]), + &(src[srcStride*i]), BLOCK_SIZE); #endif //HAVE_MMX - } + } } /** @@ -3322,23 +3322,23 @@ static inline void RENAME(duplicate)(uint8_t src[], int stride) { #ifdef HAVE_MMX - asm volatile( - "movq (%0), %%mm0 \n\t" - "add %1, %0 \n\t" - "movq %%mm0, (%0) \n\t" - "movq %%mm0, (%0, %1) \n\t" - "movq %%mm0, (%0, %1, 2) \n\t" - : "+r" (src) - : "r" ((long)-stride) - ); + asm volatile( + "movq (%0), %%mm0 \n\t" + "add %1, %0 \n\t" + "movq %%mm0, (%0) \n\t" + "movq %%mm0, (%0, %1) \n\t" + "movq %%mm0, (%0, %1, 2) \n\t" + : "+r" (src) + : "r" ((long)-stride) + ); #else - int i; - uint8_t *p=src; - for(i=0; i<3; i++) - { - p-= stride; - memcpy(p, src, 8); - } + int i; + uint8_t *p=src; + for(i=0; i<3; i++) + { + p-= stride; + memcpy(p, src, 8); + } #endif } @@ -3346,502 +3346,502 @@ * Filters array of bytes (Y or U or V values) */ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height, - QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c2) + QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c2) { - PPContext __attribute__((aligned(8))) c= *c2; //copy to stack for faster access - int x,y; + PPContext __attribute__((aligned(8))) c= *c2; //copy to stack for faster access + int x,y; #ifdef COMPILE_TIME_MODE - const int mode= COMPILE_TIME_MODE; + const int mode= COMPILE_TIME_MODE; #else - const int mode= isColor ? c.ppMode.chromMode : c.ppMode.lumMode; + const int mode= isColor ? c.ppMode.chromMode : c.ppMode.lumMode; #endif - int black=0, white=255; // blackest black and whitest white in the picture - int QPCorrecture= 256*256; - - int copyAhead; + int black=0, white=255; // blackest black and whitest white in the picture + int QPCorrecture= 256*256; + + int copyAhead; #ifdef HAVE_MMX - int i; + int i; #endif - const int qpHShift= isColor ? 4-c.hChromaSubSample : 4; - const int qpVShift= isColor ? 4-c.vChromaSubSample : 4; - - //FIXME remove - uint64_t * const yHistogram= c.yHistogram; - uint8_t * const tempSrc= srcStride > 0 ? c.tempSrc : c.tempSrc - 23*srcStride; - uint8_t * const tempDst= dstStride > 0 ? c.tempDst : c.tempDst - 23*dstStride; - //const int mbWidth= isColor ? (width+7)>>3 : (width+15)>>4; + const int qpHShift= isColor ? 4-c.hChromaSubSample : 4; + const int qpVShift= isColor ? 4-c.vChromaSubSample : 4; + + //FIXME remove + uint64_t * const yHistogram= c.yHistogram; + uint8_t * const tempSrc= srcStride > 0 ? c.tempSrc : c.tempSrc - 23*srcStride; + uint8_t * const tempDst= dstStride > 0 ? c.tempDst : c.tempDst - 23*dstStride; + //const int mbWidth= isColor ? (width+7)>>3 : (width+15)>>4; #ifdef HAVE_MMX - for(i=0; i<57; i++){ - int offset= ((i*c.ppMode.baseDcDiff)>>8) + 1; - int threshold= offset*2 + 1; - c.mmxDcOffset[i]= 0x7F - offset; - c.mmxDcThreshold[i]= 0x7F - threshold; - c.mmxDcOffset[i]*= 0x0101010101010101LL; - c.mmxDcThreshold[i]*= 0x0101010101010101LL; - } + for(i=0; i<57; i++){ + int offset= ((i*c.ppMode.baseDcDiff)>>8) + 1; + int threshold= offset*2 + 1; + c.mmxDcOffset[i]= 0x7F - offset; + c.mmxDcThreshold[i]= 0x7F - threshold; + c.mmxDcOffset[i]*= 0x0101010101010101LL; + c.mmxDcThreshold[i]*= 0x0101010101010101LL; + } #endif - if(mode & CUBIC_IPOL_DEINT_FILTER) copyAhead=16; - else if( (mode & LINEAR_BLEND_DEINT_FILTER) - || (mode & FFMPEG_DEINT_FILTER) - || (mode & LOWPASS5_DEINT_FILTER)) copyAhead=14; - else if( (mode & V_DEBLOCK) - || (mode & LINEAR_IPOL_DEINT_FILTER) - || (mode & MEDIAN_DEINT_FILTER) - || (mode & V_A_DEBLOCK)) copyAhead=13; - else if(mode & V_X1_FILTER) copyAhead=11; -// else if(mode & V_RK1_FILTER) copyAhead=10; - else if(mode & DERING) copyAhead=9; - else copyAhead=8; - - copyAhead-= 8; - - if(!isColor) - { - uint64_t sum= 0; - int i; - uint64_t maxClipped; - uint64_t clipped; - double scale; - - c.frameNum++; - // first frame is fscked so we ignore it - if(c.frameNum == 1) yHistogram[0]= width*height/64*15/256; - - for(i=0; i<256; i++) - { - sum+= yHistogram[i]; -// printf("%d ", yHistogram[i]); - } -// printf("\n\n"); - - /* we allways get a completly black picture first */ - maxClipped= (uint64_t)(sum * c.ppMode.maxClippedThreshold); - - clipped= sum; - for(black=255; black>0; black--) - { - if(clipped < maxClipped) break; - clipped-= yHistogram[black]; - } - - clipped= sum; - for(white=0; white<256; white++) - { - if(clipped < maxClipped) break; - clipped-= yHistogram[white]; - } - - scale= (double)(c.ppMode.maxAllowedY - c.ppMode.minAllowedY) / (double)(white-black); + if(mode & CUBIC_IPOL_DEINT_FILTER) copyAhead=16; + else if( (mode & LINEAR_BLEND_DEINT_FILTER) + || (mode & FFMPEG_DEINT_FILTER) + || (mode & LOWPASS5_DEINT_FILTER)) copyAhead=14; + else if( (mode & V_DEBLOCK) + || (mode & LINEAR_IPOL_DEINT_FILTER) + || (mode & MEDIAN_DEINT_FILTER) + || (mode & V_A_DEBLOCK)) copyAhead=13; + else if(mode & V_X1_FILTER) copyAhead=11; +// else if(mode & V_RK1_FILTER) copyAhead=10; + else if(mode & DERING) copyAhead=9; + else copyAhead=8; + + copyAhead-= 8; + + if(!isColor) + { + uint64_t sum= 0; + int i; + uint64_t maxClipped; + uint64_t clipped; + double scale; + + c.frameNum++; + // first frame is fscked so we ignore it + if(c.frameNum == 1) yHistogram[0]= width*height/64*15/256; + + for(i=0; i<256; i++) + { + sum+= yHistogram[i]; +// printf("%d ", yHistogram[i]); + } +// printf("\n\n"); + + /* we allways get a completly black picture first */ + maxClipped= (uint64_t)(sum * c.ppMode.maxClippedThreshold); + + clipped= sum; + for(black=255; black>0; black--) + { + if(clipped < maxClipped) break; + clipped-= yHistogram[black]; + } + + clipped= sum; + for(white=0; white<256; white++) + { + if(clipped < maxClipped) break; + clipped-= yHistogram[white]; + } + + scale= (double)(c.ppMode.maxAllowedY - c.ppMode.minAllowedY) / (double)(white-black); #ifdef HAVE_MMX2 - c.packedYScale= (uint16_t)(scale*256.0 + 0.5); - c.packedYOffset= (((black*c.packedYScale)>>8) - c.ppMode.minAllowedY) & 0xFFFF; + c.packedYScale= (uint16_t)(scale*256.0 + 0.5); + c.packedYOffset= (((black*c.packedYScale)>>8) - c.ppMode.minAllowedY) & 0xFFFF; #else - c.packedYScale= (uint16_t)(scale*1024.0 + 0.5); - c.packedYOffset= (black - c.ppMode.minAllowedY) & 0xFFFF; + c.packedYScale= (uint16_t)(scale*1024.0 + 0.5); + c.packedYOffset= (black - c.ppMode.minAllowedY) & 0xFFFF; #endif - c.packedYOffset|= c.packedYOffset<<32; - c.packedYOffset|= c.packedYOffset<<16; - - c.packedYScale|= c.packedYScale<<32; - c.packedYScale|= c.packedYScale<<16; - - if(mode & LEVEL_FIX) QPCorrecture= (int)(scale*256*256 + 0.5); - else QPCorrecture= 256*256; - } - else - { - c.packedYScale= 0x0100010001000100LL; - c.packedYOffset= 0; - QPCorrecture= 256*256; - } - - /* copy & deinterlace first row of blocks */ - y=-BLOCK_SIZE; - { - uint8_t *srcBlock= &(src[y*srcStride]); - uint8_t *dstBlock= tempDst + dstStride; - - // From this point on it is guranteed that we can read and write 16 lines downward - // finish 1 block before the next otherwise we might have a problem - // with the L1 Cache of the P4 ... or only a few blocks at a time or soemthing - for(x=0; x<width; x+=BLOCK_SIZE) - { + c.packedYOffset|= c.packedYOffset<<32; + c.packedYOffset|= c.packedYOffset<<16; + + c.packedYScale|= c.packedYScale<<32; + c.packedYScale|= c.packedYScale<<16; + + if(mode & LEVEL_FIX) QPCorrecture= (int)(scale*256*256 + 0.5); + else QPCorrecture= 256*256; + } + else + { + c.packedYScale= 0x0100010001000100LL; + c.packedYOffset= 0; + QPCorrecture= 256*256; + } + + /* copy & deinterlace first row of blocks */ + y=-BLOCK_SIZE; + { + uint8_t *srcBlock= &(src[y*srcStride]); + uint8_t *dstBlock= tempDst + dstStride; + + // From this point on it is guranteed that we can read and write 16 lines downward + // finish 1 block before the next otherwise we might have a problem + // with the L1 Cache of the P4 ... or only a few blocks at a time or soemthing + for(x=0; x<width; x+=BLOCK_SIZE) + { #ifdef HAVE_MMX2 /* - prefetchnta(srcBlock + (((x>>2)&6) + 5)*srcStride + 32); - prefetchnta(srcBlock + (((x>>2)&6) + 6)*srcStride + 32); - prefetcht0(dstBlock + (((x>>2)&6) + 5)*dstStride + 32); - prefetcht0(dstBlock + (((x>>2)&6) + 6)*dstStride + 32); + prefetchnta(srcBlock + (((x>>2)&6) + 5)*srcStride + 32); + prefetchnta(srcBlock + (((x>>2)&6) + 6)*srcStride + 32); + prefetcht0(dstBlock + (((x>>2)&6) + 5)*dstStride + 32); + prefetcht0(dstBlock + (((x>>2)&6) + 6)*dstStride + 32); */ - asm( - "mov %4, %%"REG_a" \n\t" - "shr $2, %%"REG_a" \n\t" - "and $6, %%"REG_a" \n\t" - "add %5, %%"REG_a" \n\t" - "mov %%"REG_a", %%"REG_d" \n\t" - "imul %1, %%"REG_a" \n\t" - "imul %3, %%"REG_d" \n\t" - "prefetchnta 32(%%"REG_a", %0) \n\t" - "prefetcht0 32(%%"REG_d", %2) \n\t" - "add %1, %%"REG_a" \n\t" - "add %3, %%"REG_d" \n\t" - "prefetchnta 32(%%"REG_a", %0) \n\t" - "prefetcht0 32(%%"REG_d", %2) \n\t" - :: "r" (srcBlock), "r" ((long)srcStride), "r" (dstBlock), "r" ((long)dstStride), - "g" ((long)x), "g" ((long)copyAhead) - : "%"REG_a, "%"REG_d - ); + asm( + "mov %4, %%"REG_a" \n\t" + "shr $2, %%"REG_a" \n\t" + "and $6, %%"REG_a" \n\t" + "add %5, %%"REG_a" \n\t" + "mov %%"REG_a", %%"REG_d" \n\t" + "imul %1, %%"REG_a" \n\t" + "imul %3, %%"REG_d" \n\t" + "prefetchnta 32(%%"REG_a", %0) \n\t" + "prefetcht0 32(%%"REG_d", %2) \n\t" + "add %1, %%"REG_a" \n\t" + "add %3, %%"REG_d" \n\t" + "prefetchnta 32(%%"REG_a", %0) \n\t" + "prefetcht0 32(%%"REG_d", %2) \n\t" + :: "r" (srcBlock), "r" ((long)srcStride), "r" (dstBlock), "r" ((long)dstStride), + "g" ((long)x), "g" ((long)copyAhead) + : "%"REG_a, "%"REG_d + ); #elif defined(HAVE_3DNOW) //FIXME check if this is faster on an 3dnow chip or if its faster without the prefetch or ... -/* prefetch(srcBlock + (((x>>3)&3) + 5)*srcStride + 32); - prefetch(srcBlock + (((x>>3)&3) + 9)*srcStride + 32); - prefetchw(dstBlock + (((x>>3)&3) + 5)*dstStride + 32); - prefetchw(dstBlock + (((x>>3)&3) + 9)*dstStride + 32); +/* prefetch(srcBlock + (((x>>3)&3) + 5)*srcStride + 32); + prefetch(srcBlock + (((x>>3)&3) + 9)*srcStride + 32); + prefetchw(dstBlock + (((x>>3)&3) + 5)*dstStride + 32); + prefetchw(dstBlock + (((x>>3)&3) + 9)*dstStride + 32); */ #endif - RENAME(blockCopy)(dstBlock + dstStride*8, dstStride, - srcBlock + srcStride*8, srcStride, mode & LEVEL_FIX, &c.packedYOffset); - - RENAME(duplicate)(dstBlock + dstStride*8, dstStride); - - if(mode & LINEAR_IPOL_DEINT_FILTER) - RENAME(deInterlaceInterpolateLinear)(dstBlock, dstStride); - else if(mode & LINEAR_BLEND_DEINT_FILTER) - RENAME(deInterlaceBlendLinear)(dstBlock, dstStride, c.deintTemp + x); - else if(mode & MEDIAN_DEINT_FILTER) - RENAME(deInterlaceMedian)(dstBlock, dstStride); - else if(mode & CUBIC_IPOL_DEINT_FILTER) - RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride); - else if(mode & FFMPEG_DEINT_FILTER) - RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x); - else if(mode & LOWPASS5_DEINT_FILTER) - RENAME(deInterlaceL5)(dstBlock, dstStride, c.deintTemp + x, c.deintTemp + width + x); -/* else if(mode & CUBIC_BLEND_DEINT_FILTER) - RENAME(deInterlaceBlendCubic)(dstBlock, dstStride); + RENAME(blockCopy)(dstBlock + dstStride*8, dstStride, + srcBlock + srcStride*8, srcStride, mode & LEVEL_FIX, &c.packedYOffset); + + RENAME(duplicate)(dstBlock + dstStride*8, dstStride); + + if(mode & LINEAR_IPOL_DEINT_FILTER) + RENAME(deInterlaceInterpolateLinear)(dstBlock, dstStride); + else if(mode & LINEAR_BLEND_DEINT_FILTER) + RENAME(deInterlaceBlendLinear)(dstBlock, dstStride, c.deintTemp + x); + else if(mode & MEDIAN_DEINT_FILTER) + RENAME(deInterlaceMedian)(dstBlock, dstStride); + else if(mode & CUBIC_IPOL_DEINT_FILTER) + RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride); + else if(mode & FFMPEG_DEINT_FILTER) + RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x); + else if(mode & LOWPASS5_DEINT_FILTER) + RENAME(deInterlaceL5)(dstBlock, dstStride, c.deintTemp + x, c.deintTemp + width + x); +/* else if(mode & CUBIC_BLEND_DEINT_FILTER) + RENAME(deInterlaceBlendCubic)(dstBlock, dstStride); */ - dstBlock+=8; - srcBlock+=8; - } - if(width==ABS(dstStride)) - linecpy(dst, tempDst + 9*dstStride, copyAhead, dstStride); - else - { - int i; - for(i=0; i<copyAhead; i++) - { - memcpy(dst + i*dstStride, tempDst + (9+i)*dstStride, width); - } - } - } + dstBlock+=8; + srcBlock+=8; + } + if(width==ABS(dstStride)) + linecpy(dst, tempDst + 9*dstStride, copyAhead, dstStride); + else + { + int i; + for(i=0; i<copyAhead; i++) + { + memcpy(dst + i*dstStride, tempDst + (9+i)*dstStride, width); + } + } + } //printf("\n"); - for(y=0; y<height; y+=BLOCK_SIZE) - { - //1% speedup if these are here instead of the inner loop - uint8_t *srcBlock= &(src[y*srcStride]); - uint8_t *dstBlock= &(dst[y*dstStride]); + for(y=0; y<height; y+=BLOCK_SIZE) + { + //1% speedup if these are here instead of the inner loop + uint8_t *srcBlock= &(src[y*srcStride]); + uint8_t *dstBlock= &(dst[y*dstStride]); #ifdef HAVE_MMX - uint8_t *tempBlock1= c.tempBlocks; - uint8_t *tempBlock2= c.tempBlocks + 8; + uint8_t *tempBlock1= c.tempBlocks; + uint8_t *tempBlock2= c.tempBlocks + 8; #endif - int8_t *QPptr= &QPs[(y>>qpVShift)*QPStride]; - int8_t *nonBQPptr= &c.nonBQPTable[(y>>qpVShift)*ABS(QPStride)]; - int QP=0; - /* can we mess with a 8x16 block from srcBlock/dstBlock downwards and 1 line upwards - if not than use a temporary buffer */ - if(y+15 >= height) - { - int i; - /* copy from line (copyAhead) to (copyAhead+7) of src, these will be copied with - blockcopy to dst later */ - linecpy(tempSrc + srcStride*copyAhead, srcBlock + srcStride*copyAhead, - MAX(height-y-copyAhead, 0), srcStride); - - /* duplicate last line of src to fill the void upto line (copyAhead+7) */ - for(i=MAX(height-y, 8); i<copyAhead+8; i++) - memcpy(tempSrc + srcStride*i, src + srcStride*(height-1), ABS(srcStride)); - - /* copy up to (copyAhead+1) lines of dst (line -1 to (copyAhead-1))*/ - linecpy(tempDst, dstBlock - dstStride, MIN(height-y+1, copyAhead+1), dstStride); - - /* duplicate last line of dst to fill the void upto line (copyAhead) */ - for(i=height-y+1; i<=copyAhead; i++) - memcpy(tempDst + dstStride*i, dst + dstStride*(height-1), ABS(dstStride)); - - dstBlock= tempDst + dstStride; - srcBlock= tempSrc; - } + int8_t *QPptr= &QPs[(y>>qpVShift)*QPStride]; + int8_t *nonBQPptr= &c.nonBQPTable[(y>>qpVShift)*ABS(QPStride)]; + int QP=0; + /* can we mess with a 8x16 block from srcBlock/dstBlock downwards and 1 line upwards + if not than use a temporary buffer */ + if(y+15 >= height) + { + int i; + /* copy from line (copyAhead) to (copyAhead+7) of src, these will be copied with + blockcopy to dst later */ + linecpy(tempSrc + srcStride*copyAhead, srcBlock + srcStride*copyAhead, + MAX(height-y-copyAhead, 0), srcStride); + + /* duplicate last line of src to fill the void upto line (copyAhead+7) */ + for(i=MAX(height-y, 8); i<copyAhead+8; i++) + memcpy(tempSrc + srcStride*i, src + srcStride*(height-1), ABS(srcStride)); + + /* copy up to (copyAhead+1) lines of dst (line -1 to (copyAhead-1))*/ + linecpy(tempDst, dstBlock - dstStride, MIN(height-y+1, copyAhead+1), dstStride); + + /* duplicate last line of dst to fill the void upto line (copyAhead) */ + for(i=height-y+1; i<=copyAhead; i++) + memcpy(tempDst + dstStride*i, dst + dstStride*(height-1), ABS(dstStride)); + + dstBlock= tempDst + dstStride; + srcBlock= tempSrc; + } //printf("\n"); - // From this point on it is guranteed that we can read and write 16 lines downward - // finish 1 block before the next otherwise we might have a problem - // with the L1 Cache of the P4 ... or only a few blocks at a time or soemthing - for(x=0; x<width; x+=BLOCK_SIZE) - { - const int stride= dstStride; + // From this point on it is guranteed that we can read and write 16 lines downward + // finish 1 block before the next otherwise we might have a problem + // with the L1 Cache of the P4 ... or only a few blocks at a time or soemthing + for(x=0; x<width; x+=BLOCK_SIZE) + { + const int stride= dstStride; #ifdef HAVE_MMX - uint8_t *tmpXchg; + uint8_t *tmpXchg; #endif - if(isColor) - { - QP= QPptr[x>>qpHShift]; - c.nonBQP= nonBQPptr[x>>qpHShift]; - } - else - { - QP= QPptr[x>>4]; - QP= (QP* QPCorrecture + 256*128)>>16; - c.nonBQP= nonBQPptr[x>>4]; - c.nonBQP= (c.nonBQP* QPCorrecture + 256*128)>>16; - yHistogram[ srcBlock[srcStride*12 + 4] ]++; - } - c.QP= QP; + if(isColor) + { + QP= QPptr[x>>qpHShift]; + c.nonBQP= nonBQPptr[x>>qpHShift]; + } + else + { + QP= QPptr[x>>4]; + QP= (QP* QPCorrecture + 256*128)>>16; + c.nonBQP= nonBQPptr[x>>4]; + c.nonBQP= (c.nonBQP* QPCorrecture + 256*128)>>16; + yHistogram[ srcBlock[srcStride*12 + 4] ]++; + } + c.QP= QP; #ifdef HAVE_MMX - asm volatile( - "movd %1, %%mm7 \n\t" - "packuswb %%mm7, %%mm7 \n\t" // 0, 0, 0, QP, 0, 0, 0, QP - "packuswb %%mm7, %%mm7 \n\t" // 0,QP, 0, QP, 0,QP, 0, QP - "packuswb %%mm7, %%mm7 \n\t" // QP,..., QP - "movq %%mm7, %0 \n\t" - : "=m" (c.pQPb) - : "r" (QP) - ); + asm volatile( + "movd %1, %%mm7 \n\t" + "packuswb %%mm7, %%mm7 \n\t" // 0, 0, 0, QP, 0, 0, 0, QP + "packuswb %%mm7, %%mm7 \n\t" // 0,QP, 0, QP, 0,QP, 0, QP + "packuswb %%mm7, %%mm7 \n\t" // QP,..., QP + "movq %%mm7, %0 \n\t" + : "=m" (c.pQPb) + : "r" (QP) + ); #endif #ifdef HAVE_MMX2 /* - prefetchnta(srcBlock + (((x>>2)&6) + 5)*srcStride + 32); - prefetchnta(srcBlock + (((x>>2)&6) + 6)*srcStride + 32); - prefetcht0(dstBlock + (((x>>2)&6) + 5)*dstStride + 32); - prefetcht0(dstBlock + (((x>>2)&6) + 6)*dstStride + 32); + prefetchnta(srcBlock + (((x>>2)&6) + 5)*srcStride + 32); + prefetchnta(srcBlock + (((x>>2)&6) + 6)*srcStride + 32); + prefetcht0(dstBlock + (((x>>2)&6) + 5)*dstStride + 32); + prefetcht0(dstBlock + (((x>>2)&6) + 6)*dstStride + 32); */ - asm( - "mov %4, %%"REG_a" \n\t" - "shr $2, %%"REG_a" \n\t" - "and $6, %%"REG_a" \n\t" - "add %5, %%"REG_a" \n\t" - "mov %%"REG_a", %%"REG_d" \n\t" - "imul %1, %%"REG_a" \n\t" - "imul %3, %%"REG_d" \n\t" - "prefetchnta 32(%%"REG_a", %0) \n\t" - "prefetcht0 32(%%"REG_d", %2) \n\t" - "add %1, %%"REG_a" \n\t" - "add %3, %%"REG_d" \n\t" - "prefetchnta 32(%%"REG_a", %0) \n\t" - "prefetcht0 32(%%"REG_d", %2) \n\t" - :: "r" (srcBlock), "r" ((long)srcStride), "r" (dstBlock), "r" ((long)dstStride), - "g" ((long)x), "g" ((long)copyAhead) - : "%"REG_a, "%"REG_d - ); + asm( + "mov %4, %%"REG_a" \n\t" + "shr $2, %%"REG_a" \n\t" + "and $6, %%"REG_a" \n\t" + "add %5, %%"REG_a" \n\t" + "mov %%"REG_a", %%"REG_d" \n\t" + "imul %1, %%"REG_a" \n\t" + "imul %3, %%"REG_d" \n\t" + "prefetchnta 32(%%"REG_a", %0) \n\t" + "prefetcht0 32(%%"REG_d", %2) \n\t" + "add %1, %%"REG_a" \n\t" + "add %3, %%"REG_d" \n\t" + "prefetchnta 32(%%"REG_a", %0) \n\t" + "prefetcht0 32(%%"REG_d", %2) \n\t" + :: "r" (srcBlock), "r" ((long)srcStride), "r" (dstBlock), "r" ((long)dstStride), + "g" ((long)x), "g" ((long)copyAhead) + : "%"REG_a, "%"REG_d + ); #elif defined(HAVE_3DNOW) //FIXME check if this is faster on an 3dnow chip or if its faster without the prefetch or ... -/* prefetch(srcBlock + (((x>>3)&3) + 5)*srcStride + 32); - prefetch(srcBlock + (((x>>3)&3) + 9)*srcStride + 32); - prefetchw(dstBlock + (((x>>3)&3) + 5)*dstStride + 32); - prefetchw(dstBlock + (((x>>3)&3) + 9)*dstStride + 32); +/* prefetch(srcBlock + (((x>>3)&3) + 5)*srcStride + 32); + prefetch(srcBlock + (((x>>3)&3) + 9)*srcStride + 32); + prefetchw(dstBlock + (((x>>3)&3) + 5)*dstStride + 32); + prefetchw(dstBlock + (((x>>3)&3) + 9)*dstStride + 32); */ #endif - RENAME(blockCopy)(dstBlock + dstStride*copyAhead, dstStride, - srcBlock + srcStride*copyAhead, srcStride, mode & LEVEL_FIX, &c.packedYOffset); - - if(mode & LINEAR_IPOL_DEINT_FILTER) - RENAME(deInterlaceInterpolateLinear)(dstBlock, dstStride); - else if(mode & LINEAR_BLEND_DEINT_FILTER) - RENAME(deInterlaceBlendLinear)(dstBlock, dstStride, c.deintTemp + x); - else if(mode & MEDIAN_DEINT_FILTER) - RENAME(deInterlaceMedian)(dstBlock, dstStride); - else if(mode & CUBIC_IPOL_DEINT_FILTER) - RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride); - else if(mode & FFMPEG_DEINT_FILTER) - RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x); - else if(mode & LOWPASS5_DEINT_FILTER) - RENAME(deInterlaceL5)(dstBlock, dstStride, c.deintTemp + x, c.deintTemp + width + x); -/* else if(mode & CUBIC_BLEND_DEINT_FILTER) - RENAME(deInterlaceBlendCubic)(dstBlock, dstStride); + RENAME(blockCopy)(dstBlock + dstStride*copyAhead, dstStride, + srcBlock + srcStride*copyAhead, srcStride, mode & LEVEL_FIX, &c.packedYOffset); + + if(mode & LINEAR_IPOL_DEINT_FILTER) + RENAME(deInterlaceInterpolateLinear)(dstBlock, dstStride); + else if(mode & LINEAR_BLEND_DEINT_FILTER) + RENAME(deInterlaceBlendLinear)(dstBlock, dstStride, c.deintTemp + x); + else if(mode & MEDIAN_DEINT_FILTER) + RENAME(deInterlaceMedian)(dstBlock, dstStride); + else if(mode & CUBIC_IPOL_DEINT_FILTER) + RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride); + else if(mode & FFMPEG_DEINT_FILTER) + RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x); + else if(mode & LOWPASS5_DEINT_FILTER) + RENAME(deInterlaceL5)(dstBlock, dstStride, c.deintTemp + x, c.deintTemp + width + x); +/* else if(mode & CUBIC_BLEND_DEINT_FILTER) + RENAME(deInterlaceBlendCubic)(dstBlock, dstStride); */ - /* only deblock if we have 2 blocks */ - if(y + 8 < height) - { - if(mode & V_X1_FILTER) - RENAME(vertX1Filter)(dstBlock, stride, &c); - else if(mode & V_DEBLOCK) - { - const int t= RENAME(vertClassify)(dstBlock, stride, &c); - - if(t==1) - RENAME(doVertLowPass)(dstBlock, stride, &c); - else if(t==2) - RENAME(doVertDefFilter)(dstBlock, stride, &c); - }else if(mode & V_A_DEBLOCK){ - RENAME(do_a_deblock)(dstBlock, stride, 1, &c); - } - } + /* only deblock if we have 2 blocks */ + if(y + 8 < height) + { + if(mode & V_X1_FILTER) + RENAME(vertX1Filter)(dstBlock, stride, &c); + else if(mode & V_DEBLOCK) + { + const int t= RENAME(vertClassify)(dstBlock, stride, &c); + + if(t==1) + RENAME(doVertLowPass)(dstBlock, stride, &c); + else if(t==2) + RENAME(doVertDefFilter)(dstBlock, stride, &c); + }else if(mode & V_A_DEBLOCK){ + RENAME(do_a_deblock)(dstBlock, stride, 1, &c); + } + } #ifdef HAVE_MMX - RENAME(transpose1)(tempBlock1, tempBlock2, dstBlock, dstStride); + RENAME(transpose1)(tempBlock1, tempBlock2, dstBlock, dstStride); #endif - /* check if we have a previous block to deblock it with dstBlock */ - if(x - 8 >= 0) - { + /* check if we have a previous block to deblock it with dstBlock */ + if(x - 8 >= 0) + { #ifdef HAVE_MMX - if(mode & H_X1_FILTER) - RENAME(vertX1Filter)(tempBlock1, 16, &c); - else if(mode & H_DEBLOCK) - { + if(mode & H_X1_FILTER) + RENAME(vertX1Filter)(tempBlock1, 16, &c); + else if(mode & H_DEBLOCK) + { //START_TIMER - const int t= RENAME(vertClassify)(tempBlock1, 16, &c); + const int t= RENAME(vertClassify)(tempBlock1, 16, &c); //STOP_TIMER("dc & minmax") if(t==1) - RENAME(doVertLowPass)(tempBlock1, 16, &c); - else if(t==2) - RENAME(doVertDefFilter)(tempBlock1, 16, &c); - }else if(mode & H_A_DEBLOCK){ - RENAME(do_a_deblock)(tempBlock1, 16, 1, &c); - } - - RENAME(transpose2)(dstBlock-4, dstStride, tempBlock1 + 4*16); + RENAME(doVertLowPass)(tempBlock1, 16, &c); + else if(t==2) + RENAME(doVertDefFilter)(tempBlock1, 16, &c); + }else if(mode & H_A_DEBLOCK){ + RENAME(do_a_deblock)(tempBlock1, 16, 1, &c); + } + + RENAME(transpose2)(dstBlock-4, dstStride, tempBlock1 + 4*16); #else - if(mode & H_X1_FILTER) - horizX1Filter(dstBlock-4, stride, QP); - else if(mode & H_DEBLOCK) - { + if(mode & H_X1_FILTER) + horizX1Filter(dstBlock-4, stride, QP); + else if(mode & H_DEBLOCK) + { #ifdef HAVE_ALTIVEC - unsigned char __attribute__ ((aligned(16))) tempBlock[272]; - transpose_16x8_char_toPackedAlign_altivec(tempBlock, dstBlock - (4 + 1), stride); - - const int t=vertClassify_altivec(tempBlock-48, 16, &c); - if(t==1) { - doVertLowPass_altivec(tempBlock-48, 16, &c); + unsigned char __attribute__ ((aligned(16))) tempBlock[272]; + transpose_16x8_char_toPackedAlign_altivec(tempBlock, dstBlock - (4 + 1), stride); + + const int t=vertClassify_altivec(tempBlock-48, 16, &c); + if(t==1) { + doVertLowPass_altivec(tempBlock-48, 16, &c); transpose_8x16_char_fromPackedAlign_altivec(dstBlock - (4 + 1), tempBlock, stride); } - else if(t==2) { - doVertDefFilter_altivec(tempBlock-48, 16, &c); + else if(t==2) { + doVertDefFilter_altivec(tempBlock-48, 16, &c); transpose_8x16_char_fromPackedAlign_altivec(dstBlock - (4 + 1), tempBlock, stride); } #else - const int t= RENAME(horizClassify)(dstBlock-4, stride, &c); - - if(t==1) - RENAME(doHorizLowPass)(dstBlock-4, stride, &c); - else if(t==2) - RENAME(doHorizDefFilter)(dstBlock-4, stride, &c); + const int t= RENAME(horizClassify)(dstBlock-4, stride, &c); + + if(t==1) + RENAME(doHorizLowPass)(dstBlock-4, stride, &c); + else if(t==2) + RENAME(doHorizDefFilter)(dstBlock-4, stride, &c); #endif - }else if(mode & H_A_DEBLOCK){ - RENAME(do_a_deblock)(dstBlock-8, 1, stride, &c); - } + }else if(mode & H_A_DEBLOCK){ + RENAME(do_a_deblock)(dstBlock-8, 1, stride, &c); + } #endif //HAVE_MMX - if(mode & DERING) - { - //FIXME filter first line - if(y>0) RENAME(dering)(dstBlock - stride - 8, stride, &c); - } - - if(mode & TEMP_NOISE_FILTER) - { - RENAME(tempNoiseReducer)(dstBlock-8, stride, - c.tempBlured[isColor] + y*dstStride + x, - c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3), - c.ppMode.maxTmpNoise); - } - } - - dstBlock+=8; - srcBlock+=8; + if(mode & DERING) + { + //FIXME filter first line + if(y>0) RENAME(dering)(dstBlock - stride - 8, stride, &c); + } + + if(mode & TEMP_NOISE_FILTER) + { + RENAME(tempNoiseReducer)(dstBlock-8, stride, + c.tempBlured[isColor] + y*dstStride + x, + c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3), + c.ppMode.maxTmpNoise); + } + } + + dstBlock+=8; + srcBlock+=8; #ifdef HAVE_MMX - tmpXchg= tempBlock1; - tempBlock1= tempBlock2; - tempBlock2 = tmpXchg; + tmpXchg= tempBlock1; + tempBlock1= tempBlock2; + tempBlock2 = tmpXchg; #endif - } - - if(mode & DERING) - { - if(y > 0) RENAME(dering)(dstBlock - dstStride - 8, dstStride, &c); - } - - if((mode & TEMP_NOISE_FILTER)) - { - RENAME(tempNoiseReducer)(dstBlock-8, dstStride, - c.tempBlured[isColor] + y*dstStride + x, - c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3), - c.ppMode.maxTmpNoise); - } - - /* did we use a tmp buffer for the last lines*/ - if(y+15 >= height) - { - uint8_t *dstBlock= &(dst[y*dstStride]); - if(width==ABS(dstStride)) - linecpy(dstBlock, tempDst + dstStride, height-y, dstStride); - else - { - int i; - for(i=0; i<height-y; i++) - { - memcpy(dstBlock + i*dstStride, tempDst + (i+1)*dstStride, width); - } - } - } + } + + if(mode & DERING) + { + if(y > 0) RENAME(dering)(dstBlock - dstStride - 8, dstStride, &c); + } + + if((mode & TEMP_NOISE_FILTER)) + { + RENAME(tempNoiseReducer)(dstBlock-8, dstStride, + c.tempBlured[isColor] + y*dstStride + x, + c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3), + c.ppMode.maxTmpNoise); + } + + /* did we use a tmp buffer for the last lines*/ + if(y+15 >= height) + { + uint8_t *dstBlock= &(dst[y*dstStride]); + if(width==ABS(dstStride)) + linecpy(dstBlock, tempDst + dstStride, height-y, dstStride); + else + { + int i; + for(i=0; i<height-y; i++) + { + memcpy(dstBlock + i*dstStride, tempDst + (i+1)*dstStride, width); + } + } + } /* - for(x=0; x<width; x+=32) - { - volatile int i; - i+= + dstBlock[x + 7*dstStride] + dstBlock[x + 8*dstStride] - + dstBlock[x + 9*dstStride] + dstBlock[x +10*dstStride] - + dstBlock[x +11*dstStride] + dstBlock[x +12*dstStride]; -// + dstBlock[x +13*dstStride] -// + dstBlock[x +14*dstStride] + dstBlock[x +15*dstStride]; - }*/ - } + for(x=0; x<width; x+=32) + { + volatile int i; + i+= + dstBlock[x + 7*dstStride] + dstBlock[x + 8*dstStride] + + dstBlock[x + 9*dstStride] + dstBlock[x +10*dstStride] + + dstBlock[x +11*dstStride] + dstBlock[x +12*dstStride]; +// + dstBlock[x +13*dstStride] +// + dstBlock[x +14*dstStride] + dstBlock[x +15*dstStride]; + }*/ + } #ifdef HAVE_3DNOW - asm volatile("femms"); + asm volatile("femms"); #elif defined (HAVE_MMX) - asm volatile("emms"); + asm volatile("emms"); #endif #ifdef DEBUG_BRIGHTNESS - if(!isColor) - { - int max=1; - int i; - for(i=0; i<256; i++) - if(yHistogram[i] > max) max=yHistogram[i]; - - for(i=1; i<256; i++) - { - int x; - int start=yHistogram[i-1]/(max/256+1); - int end=yHistogram[i]/(max/256+1); - int inc= end > start ? 1 : -1; - for(x=start; x!=end+inc; x+=inc) - dst[ i*dstStride + x]+=128; - } - - for(i=0; i<100; i+=2) - { - dst[ (white)*dstStride + i]+=128; - dst[ (black)*dstStride + i]+=128; - } - - } + if(!isColor) + { + int max=1; + int i; + for(i=0; i<256; i++) + if(yHistogram[i] > max) max=yHistogram[i]; + + for(i=1; i<256; i++) + { + int x; + int start=yHistogram[i-1]/(max/256+1); + int end=yHistogram[i]/(max/256+1); + int inc= end > start ? 1 : -1; + for(x=start; x!=end+inc; x+=inc) + dst[ i*dstStride + x]+=128; + } + + for(i=0; i<100; i+=2) + { + dst[ (white)*dstStride + i]+=128; + dst[ (black)*dstStride + i]+=128; + } + + } #endif - *c2= c; //copy local context back + *c2= c; //copy local context back }
--- a/mem.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mem.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,8 +64,8 @@ Indeed, we should align it: on 4 for 386 on 16 for 486 - on 32 for 586, PPro - k6-III - on 64 for K7 (maybe for P3 too). + on 32 for 586, PPro - k6-III + on 64 for K7 (maybe for P3 too). Because L1 and L2 caches are aligned on those values. But I don't want to code such logic here! */ @@ -76,13 +76,13 @@ Why not larger? because i didnt see a difference in benchmarks ... */ /* benchmarks with p3 - memalign(64)+1 3071,3051,3032 - memalign(64)+2 3051,3032,3041 - memalign(64)+4 2911,2896,2915 - memalign(64)+8 2545,2554,2550 - memalign(64)+16 2543,2572,2563 - memalign(64)+32 2546,2545,2571 - memalign(64)+64 2570,2533,2558 + memalign(64)+1 3071,3051,3032 + memalign(64)+2 3051,3032,3041 + memalign(64)+4 2911,2896,2915 + memalign(64)+8 2545,2554,2550 + memalign(64)+16 2543,2572,2563 + memalign(64)+32 2546,2545,2571 + memalign(64)+64 2570,2533,2558 btw, malloc seems to do 8 byte alignment by default here */
--- a/mjpeg.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mjpeg.c Thu Dec 22 01:10:11 2005 +0000 @@ -54,26 +54,26 @@ /* JPEG marker codes */ typedef enum { /* start of frame */ - SOF0 = 0xc0, /* baseline */ - SOF1 = 0xc1, /* extended sequential, huffman */ - SOF2 = 0xc2, /* progressive, huffman */ - SOF3 = 0xc3, /* lossless, huffman */ + SOF0 = 0xc0, /* baseline */ + SOF1 = 0xc1, /* extended sequential, huffman */ + SOF2 = 0xc2, /* progressive, huffman */ + SOF3 = 0xc3, /* lossless, huffman */ - SOF5 = 0xc5, /* differential sequential, huffman */ - SOF6 = 0xc6, /* differential progressive, huffman */ - SOF7 = 0xc7, /* differential lossless, huffman */ - JPG = 0xc8, /* reserved for JPEG extension */ - SOF9 = 0xc9, /* extended sequential, arithmetic */ - SOF10 = 0xca, /* progressive, arithmetic */ - SOF11 = 0xcb, /* lossless, arithmetic */ + SOF5 = 0xc5, /* differential sequential, huffman */ + SOF6 = 0xc6, /* differential progressive, huffman */ + SOF7 = 0xc7, /* differential lossless, huffman */ + JPG = 0xc8, /* reserved for JPEG extension */ + SOF9 = 0xc9, /* extended sequential, arithmetic */ + SOF10 = 0xca, /* progressive, arithmetic */ + SOF11 = 0xcb, /* lossless, arithmetic */ - SOF13 = 0xcd, /* differential sequential, arithmetic */ - SOF14 = 0xce, /* differential progressive, arithmetic */ - SOF15 = 0xcf, /* differential lossless, arithmetic */ + SOF13 = 0xcd, /* differential sequential, arithmetic */ + SOF14 = 0xce, /* differential progressive, arithmetic */ + SOF15 = 0xcf, /* differential lossless, arithmetic */ - DHT = 0xc4, /* define huffman tables */ + DHT = 0xc4, /* define huffman tables */ - DAC = 0xcc, /* define arithmetic-coding conditioning */ + DAC = 0xcc, /* define arithmetic-coding conditioning */ /* restart with modulo 8 count "m" */ RST0 = 0xd0, @@ -85,14 +85,14 @@ RST6 = 0xd6, RST7 = 0xd7, - SOI = 0xd8, /* start of image */ - EOI = 0xd9, /* end of image */ - SOS = 0xda, /* start of scan */ - DQT = 0xdb, /* define quantization tables */ - DNL = 0xdc, /* define number of lines */ - DRI = 0xdd, /* define restart interval */ - DHP = 0xde, /* define hierarchical progression */ - EXP = 0xdf, /* expand reference components */ + SOI = 0xd8, /* start of image */ + EOI = 0xd9, /* end of image */ + SOS = 0xda, /* start of scan */ + DQT = 0xdb, /* define quantization tables */ + DNL = 0xdc, /* define number of lines */ + DRI = 0xdd, /* define restart interval */ + DHP = 0xde, /* define hierarchical progression */ + EXP = 0xdf, /* expand reference components */ APP0 = 0xe0, APP1 = 0xe1, @@ -118,17 +118,17 @@ JPG4 = 0xf4, JPG5 = 0xf5, JPG6 = 0xf6, - SOF48 = 0xf7, ///< JPEG-LS - LSE = 0xf8, ///< JPEG-LS extension parameters + SOF48 = 0xf7, ///< JPEG-LS + LSE = 0xf8, ///< JPEG-LS extension parameters JPG9 = 0xf9, JPG10 = 0xfa, JPG11 = 0xfb, JPG12 = 0xfc, JPG13 = 0xfd, - COM = 0xfe, /* comment */ + COM = 0xfe, /* comment */ - TEM = 0x01, /* temporary private use for arithmetic coding */ + TEM = 0x01, /* temporary private use for arithmetic coding */ /* 0x02 -> 0xbf reserved */ } JPEG_MARKER; @@ -583,7 +583,7 @@ } static inline void mjpeg_encode_dc(MpegEncContext *s, int val, - uint8_t *huff_size, uint16_t *huff_code) + uint8_t *huff_size, uint16_t *huff_code) { int mant, nbits; @@ -935,10 +935,10 @@ if (avctx->flags & CODEC_FLAG_EXTERN_HUFF) { - av_log(avctx, AV_LOG_INFO, "mjpeg: using external huffman table\n"); - init_get_bits(&s->gb, avctx->extradata, avctx->extradata_size*8); - mjpeg_decode_dht(s); - /* should check for error - but dunno */ + av_log(avctx, AV_LOG_INFO, "mjpeg: using external huffman table\n"); + init_get_bits(&s->gb, avctx->extradata, avctx->extradata_size*8); + mjpeg_decode_dht(s); + /* should check for error - but dunno */ } return 0; @@ -1017,10 +1017,10 @@ while (len >= 65) { /* only 8 bit precision handled */ if (get_bits(&s->gb, 4) != 0) - { - dprintf("dqt: 16bit precision\n"); + { + dprintf("dqt: 16bit precision\n"); return -1; - } + } index = get_bits(&s->gb, 4); if (index >= 4) return -1; @@ -1028,14 +1028,14 @@ /* read quant table */ for(i=0;i<64;i++) { j = s->scantable.permutated[i]; - s->quant_matrixes[index][j] = get_bits(&s->gb, 8); + s->quant_matrixes[index][j] = get_bits(&s->gb, 8); } //XXX FIXME finetune, and perhaps add dc too s->qscale[index]= FFMAX( s->quant_matrixes[index][s->scantable.permutated[1]], s->quant_matrixes[index][s->scantable.permutated[8]]) >> 1; - dprintf("qscale[%d]: %d\n", index, s->qscale[index]); + dprintf("qscale[%d]: %d\n", index, s->qscale[index]); len -= 65; } @@ -1132,7 +1132,7 @@ if (s->quant_index[i] >= 4) return -1; dprintf("component %d %d:%d id: %d quant:%d\n", i, s->h_count[i], - s->v_count[i], s->component_id[i], s->quant_index[i]); + s->v_count[i], s->component_id[i], s->quant_index[i]); } if(s->v_max==1 && s->h_max==1 && s->lossless==1) s->rgb=1; @@ -1151,7 +1151,7 @@ s->org_height != 0 && s->height < ((s->org_height * 3) / 4)) { s->interlaced = 1; -// s->bottom_field = (s->interlace_polarity) ? 1 : 0; +// s->bottom_field = (s->interlace_polarity) ? 1 : 0; s->bottom_field = 0; s->avctx->height *= 2; } @@ -1202,7 +1202,7 @@ if (len != (8+(3*nb_components))) { - dprintf("decode_sof0: error, len(%d) mismatch\n", len); + dprintf("decode_sof0: error, len(%d) mismatch\n", len); } return 0; @@ -1214,7 +1214,7 @@ code = get_vlc2(&s->gb, s->vlcs[0][dc_index].table, 9, 2); if (code < 0) { - dprintf("mjpeg_decode_dc: bad vlc: %d:%d (%p)\n", 0, dc_index, + dprintf("mjpeg_decode_dc: bad vlc: %d:%d (%p)\n", 0, dc_index, &s->vlcs[0][dc_index]); return 0xffff; } @@ -1247,7 +1247,7 @@ ac_vlc = &s->vlcs[1][ac_index]; i = 1; for(;;) { - code = get_vlc2(&s->gb, s->vlcs[1][ac_index].table, 9, 2); + code = get_vlc2(&s->gb, s->vlcs[1][ac_index].table, 9, 2); if (code < 0) { dprintf("error ac\n"); @@ -1452,7 +1452,7 @@ dprintf("error y=%d x=%d\n", mb_y, mb_x); return -1; } -// dprintf("mb: %d %d processed\n", mb_y, mb_x); +// dprintf("mb: %d %d processed\n", mb_y, mb_x); ptr = s->picture.data[c] + (((s->linesize[c] * (v * mb_y + y) * 8) + (h * mb_x + x) * 8) >> s->avctx->lowres); @@ -1491,29 +1491,29 @@ nb_components = get_bits(&s->gb, 8); if (len != 6+2*nb_components) { - dprintf("decode_sos: invalid len (%d)\n", len); - return -1; + dprintf("decode_sos: invalid len (%d)\n", len); + return -1; } /* XXX: only interleaved scan accepted */ if (nb_components != s->nb_components) { - dprintf("decode_sos: components(%d) mismatch\n", nb_components); + dprintf("decode_sos: components(%d) mismatch\n", nb_components); return -1; } vmax = 0; hmax = 0; for(i=0;i<nb_components;i++) { id = get_bits(&s->gb, 8) - 1; - dprintf("component: %d\n", id); + dprintf("component: %d\n", id); /* find component index */ for(index=0;index<s->nb_components;index++) if (id == s->component_id[index]) break; if (index == s->nb_components) - { - dprintf("decode_sos: index(%d) out of components\n", index); + { + dprintf("decode_sos: index(%d) out of components\n", index); return -1; - } + } s->comp_index[i] = index; @@ -1524,26 +1524,26 @@ s->dc_index[i] = get_bits(&s->gb, 4); s->ac_index[i] = get_bits(&s->gb, 4); - if (s->dc_index[i] < 0 || s->ac_index[i] < 0 || - s->dc_index[i] >= 4 || s->ac_index[i] >= 4) - goto out_of_range; + if (s->dc_index[i] < 0 || s->ac_index[i] < 0 || + s->dc_index[i] >= 4 || s->ac_index[i] >= 4) + goto out_of_range; #if 0 //buggy - switch(s->start_code) - { - case SOF0: - if (dc_index[i] > 1 || ac_index[i] > 1) - goto out_of_range; - break; - case SOF1: - case SOF2: - if (dc_index[i] > 3 || ac_index[i] > 3) - goto out_of_range; - break; - case SOF3: - if (dc_index[i] > 3 || ac_index[i] != 0) - goto out_of_range; - break; - } + switch(s->start_code) + { + case SOF0: + if (dc_index[i] > 1 || ac_index[i] > 1) + goto out_of_range; + break; + case SOF1: + case SOF2: + if (dc_index[i] > 3 || ac_index[i] > 3) + goto out_of_range; + break; + case SOF3: + if (dc_index[i] > 3 || ac_index[i] != 0) + goto out_of_range; + break; + } #endif } @@ -1605,7 +1605,7 @@ static int mjpeg_decode_dri(MJpegDecodeContext *s) { if (get_bits(&s->gb, 16) != 4) - return -1; + return -1; s->restart_interval = get_bits(&s->gb, 16); s->restart_count = 0; dprintf("restart interval: %d\n", s->restart_interval); @@ -1619,7 +1619,7 @@ len = get_bits(&s->gb, 16); if (len < 5) - return -1; + return -1; if(8*len + get_bits_count(&s->gb) > s->gb.size_in_bits) return -1; @@ -1636,35 +1636,35 @@ informations, but it's always present in AVID creates files */ if (id == ff_get_fourcc("AVI1")) { - /* structure: - 4bytes AVI1 - 1bytes polarity - 1bytes always zero - 4bytes field_size - 4bytes field_size_less_padding - */ - s->buggy_avid = 1; -// if (s->first_picture) -// printf("mjpeg: workarounding buggy AVID\n"); - s->interlace_polarity = get_bits(&s->gb, 8); + /* structure: + 4bytes AVI1 + 1bytes polarity + 1bytes always zero + 4bytes field_size + 4bytes field_size_less_padding + */ + s->buggy_avid = 1; +// if (s->first_picture) +// printf("mjpeg: workarounding buggy AVID\n"); + s->interlace_polarity = get_bits(&s->gb, 8); #if 0 - skip_bits(&s->gb, 8); - skip_bits(&s->gb, 32); - skip_bits(&s->gb, 32); - len -= 10; + skip_bits(&s->gb, 8); + skip_bits(&s->gb, 32); + skip_bits(&s->gb, 32); + len -= 10; #endif -// if (s->interlace_polarity) -// printf("mjpeg: interlace polarity: %d\n", s->interlace_polarity); - goto out; +// if (s->interlace_polarity) +// printf("mjpeg: interlace polarity: %d\n", s->interlace_polarity); + goto out; } // len -= 2; if (id == ff_get_fourcc("JFIF")) { - int t_w, t_h, v1, v2; - skip_bits(&s->gb, 8); /* the trailing zero-byte */ - v1= get_bits(&s->gb, 8); + int t_w, t_h, v1, v2; + skip_bits(&s->gb, 8); /* the trailing zero-byte */ + v1= get_bits(&s->gb, 8); v2= get_bits(&s->gb, 8); skip_bits(&s->gb, 8); @@ -1678,37 +1678,37 @@ s->avctx->sample_aspect_ratio.den ); - t_w = get_bits(&s->gb, 8); - t_h = get_bits(&s->gb, 8); - if (t_w && t_h) - { - /* skip thumbnail */ - if (len-10-(t_w*t_h*3) > 0) - len -= t_w*t_h*3; - } - len -= 10; - goto out; + t_w = get_bits(&s->gb, 8); + t_h = get_bits(&s->gb, 8); + if (t_w && t_h) + { + /* skip thumbnail */ + if (len-10-(t_w*t_h*3) > 0) + len -= t_w*t_h*3; + } + len -= 10; + goto out; } if (id == ff_get_fourcc("Adob") && (get_bits(&s->gb, 8) == 'e')) { if (s->avctx->debug & FF_DEBUG_PICT_INFO) av_log(s->avctx, AV_LOG_INFO, "mjpeg: Adobe header found\n"); - skip_bits(&s->gb, 16); /* version */ - skip_bits(&s->gb, 16); /* flags0 */ - skip_bits(&s->gb, 16); /* flags1 */ - skip_bits(&s->gb, 8); /* transform */ - len -= 7; - goto out; + skip_bits(&s->gb, 16); /* version */ + skip_bits(&s->gb, 16); /* flags0 */ + skip_bits(&s->gb, 16); /* flags1 */ + skip_bits(&s->gb, 8); /* transform */ + len -= 7; + goto out; } if (id == ff_get_fourcc("LJIF")){ if (s->avctx->debug & FF_DEBUG_PICT_INFO) av_log(s->avctx, AV_LOG_INFO, "Pegasus lossless jpeg header found\n"); - skip_bits(&s->gb, 16); /* version ? */ - skip_bits(&s->gb, 16); /* unknwon always 0? */ - skip_bits(&s->gb, 16); /* unknwon always 0? */ - skip_bits(&s->gb, 16); /* unknwon always 0? */ + skip_bits(&s->gb, 16); /* version ? */ + skip_bits(&s->gb, 16); /* unknwon always 0? */ + skip_bits(&s->gb, 16); /* unknwon always 0? */ + skip_bits(&s->gb, 16); /* unknwon always 0? */ switch( get_bits(&s->gb, 8)){ case 1: s->rgb= 1; @@ -1728,32 +1728,32 @@ /* Apple MJPEG-A */ if ((s->start_code == APP1) && (len > (0x28 - 8))) { - id = (get_bits(&s->gb, 16) << 16) | get_bits(&s->gb, 16); - id = be2me_32(id); - len -= 4; - if (id == ff_get_fourcc("mjpg")) /* Apple MJPEG-A */ - { + id = (get_bits(&s->gb, 16) << 16) | get_bits(&s->gb, 16); + id = be2me_32(id); + len -= 4; + if (id == ff_get_fourcc("mjpg")) /* Apple MJPEG-A */ + { #if 0 - skip_bits(&s->gb, 32); /* field size */ - skip_bits(&s->gb, 32); /* pad field size */ - skip_bits(&s->gb, 32); /* next off */ - skip_bits(&s->gb, 32); /* quant off */ - skip_bits(&s->gb, 32); /* huff off */ - skip_bits(&s->gb, 32); /* image off */ - skip_bits(&s->gb, 32); /* scan off */ - skip_bits(&s->gb, 32); /* data off */ + skip_bits(&s->gb, 32); /* field size */ + skip_bits(&s->gb, 32); /* pad field size */ + skip_bits(&s->gb, 32); /* next off */ + skip_bits(&s->gb, 32); /* quant off */ + skip_bits(&s->gb, 32); /* huff off */ + skip_bits(&s->gb, 32); /* image off */ + skip_bits(&s->gb, 32); /* scan off */ + skip_bits(&s->gb, 32); /* data off */ #endif if (s->avctx->debug & FF_DEBUG_PICT_INFO) - av_log(s->avctx, AV_LOG_INFO, "mjpeg: Apple MJPEG-A header found\n"); - } + av_log(s->avctx, AV_LOG_INFO, "mjpeg: Apple MJPEG-A header found\n"); + } } out: /* slow but needed for extreme adobe jpegs */ if (len < 0) - av_log(s->avctx, AV_LOG_ERROR, "mjpeg: error, decode_app parser read over the end\n"); + av_log(s->avctx, AV_LOG_ERROR, "mjpeg: error, decode_app parser read over the end\n"); while(--len > 0) - skip_bits(&s->gb, 8); + skip_bits(&s->gb, 8); return 0; } @@ -1762,32 +1762,32 @@ { int len = get_bits(&s->gb, 16); if (len >= 2 && 8*len - 16 + get_bits_count(&s->gb) <= s->gb.size_in_bits) { - uint8_t *cbuf = av_malloc(len - 1); - if (cbuf) { - int i; - for (i = 0; i < len - 2; i++) - cbuf[i] = get_bits(&s->gb, 8); - if (i > 0 && cbuf[i-1] == '\n') - cbuf[i-1] = 0; - else - cbuf[i] = 0; + uint8_t *cbuf = av_malloc(len - 1); + if (cbuf) { + int i; + for (i = 0; i < len - 2; i++) + cbuf[i] = get_bits(&s->gb, 8); + if (i > 0 && cbuf[i-1] == '\n') + cbuf[i-1] = 0; + else + cbuf[i] = 0; if(s->avctx->debug & FF_DEBUG_PICT_INFO) av_log(s->avctx, AV_LOG_INFO, "mjpeg comment: '%s'\n", cbuf); - /* buggy avid, it puts EOI only at every 10th frame */ - if (!strcmp(cbuf, "AVID")) - { - s->buggy_avid = 1; - // if (s->first_picture) - // printf("mjpeg: workarounding buggy AVID\n"); - } + /* buggy avid, it puts EOI only at every 10th frame */ + if (!strcmp(cbuf, "AVID")) + { + s->buggy_avid = 1; + // if (s->first_picture) + // printf("mjpeg: workarounding buggy AVID\n"); + } else if(!strcmp(cbuf, "CS=ITU601")){ s->cs_itu601= 1; } - av_free(cbuf); - } + av_free(cbuf); + } } return 0; @@ -1830,13 +1830,13 @@ buf_ptr = *pbuf_ptr; while (buf_ptr < buf_end) { v = *buf_ptr++; - v2 = *buf_ptr; + v2 = *buf_ptr; if ((v == 0xff) && (v2 >= 0xc0) && (v2 <= 0xfe) && buf_ptr < buf_end) { - val = *buf_ptr++; - goto found; + val = *buf_ptr++; + goto found; } #ifdef DEBUG - skipped++; + skipped++; #endif } val = -1; @@ -1862,74 +1862,74 @@ while (buf_ptr < buf_end) { /* find start next marker */ start_code = find_marker(&buf_ptr, buf_end); - { - /* EOF */ + { + /* EOF */ if (start_code < 0) { - goto the_end; + goto the_end; } else { dprintf("marker=%x avail_size_in_buf=%d\n", start_code, buf_end - buf_ptr); - if ((buf_end - buf_ptr) > s->buffer_size) - { - av_free(s->buffer); - s->buffer_size = buf_end-buf_ptr; + if ((buf_end - buf_ptr) > s->buffer_size) + { + av_free(s->buffer); + s->buffer_size = buf_end-buf_ptr; s->buffer = av_malloc(s->buffer_size + FF_INPUT_BUFFER_PADDING_SIZE); - dprintf("buffer too small, expanding to %d bytes\n", - s->buffer_size); - } + dprintf("buffer too small, expanding to %d bytes\n", + s->buffer_size); + } - /* unescape buffer of SOS */ - if (start_code == SOS) - { - uint8_t *src = buf_ptr; - uint8_t *dst = s->buffer; + /* unescape buffer of SOS */ + if (start_code == SOS) + { + uint8_t *src = buf_ptr; + uint8_t *dst = s->buffer; - while (src<buf_end) - { - uint8_t x = *(src++); + while (src<buf_end) + { + uint8_t x = *(src++); - *(dst++) = x; - if (x == 0xff) - { + *(dst++) = x; + if (x == 0xff) + { while(src<buf_end && x == 0xff) x = *(src++); - if (x >= 0xd0 && x <= 0xd7) - *(dst++) = x; - else if (x) - break; - } - } - init_get_bits(&s->gb, s->buffer, (dst - s->buffer)*8); + if (x >= 0xd0 && x <= 0xd7) + *(dst++) = x; + else if (x) + break; + } + } + init_get_bits(&s->gb, s->buffer, (dst - s->buffer)*8); - dprintf("escaping removed %d bytes\n", - (buf_end - buf_ptr) - (dst - s->buffer)); - } - else - init_get_bits(&s->gb, buf_ptr, (buf_end - buf_ptr)*8); + dprintf("escaping removed %d bytes\n", + (buf_end - buf_ptr) - (dst - s->buffer)); + } + else + init_get_bits(&s->gb, buf_ptr, (buf_end - buf_ptr)*8); - s->start_code = start_code; + s->start_code = start_code; if(s->avctx->debug & FF_DEBUG_STARTCODE){ av_log(s->avctx, AV_LOG_DEBUG, "startcode: %X\n", start_code); } - /* process markers */ - if (start_code >= 0xd0 && start_code <= 0xd7) { - dprintf("restart marker: %d\n", start_code&0x0f); - /* APP fields */ - } else if (start_code >= APP0 && start_code <= APP15) { - mjpeg_decode_app(s); - /* Comment */ - } else if (start_code == COM){ - mjpeg_decode_com(s); - } + /* process markers */ + if (start_code >= 0xd0 && start_code <= 0xd7) { + dprintf("restart marker: %d\n", start_code&0x0f); + /* APP fields */ + } else if (start_code >= APP0 && start_code <= APP15) { + mjpeg_decode_app(s); + /* Comment */ + } else if (start_code == COM){ + mjpeg_decode_com(s); + } switch(start_code) { case SOI: - s->restart_interval = 0; + s->restart_interval = 0; reset_ls_coding_parameters(s, 1); - s->restart_count = 0; + s->restart_count = 0; /* nothing to do on SOI */ break; case DQT: @@ -1944,12 +1944,12 @@ case SOF0: s->lossless=0; if (mjpeg_decode_sof(s) < 0) - return -1; + return -1; break; case SOF3: s->lossless=1; if (mjpeg_decode_sof(s) < 0) - return -1; + return -1; break; case SOF48: s->lossless=1; @@ -1961,11 +1961,11 @@ if (decode_lse(s) < 0) return -1; break; - case EOI: - if ((s->buggy_avid && !s->interlaced) || s->restart_interval) + case EOI: + if ((s->buggy_avid && !s->interlaced) || s->restart_interval) break; eoi_parser: - { + { if (s->interlaced) { s->bottom_field ^= 1; /* if not bottom field, do not output image yet */ @@ -1987,41 +1987,41 @@ goto the_end; } - break; + break; case SOS: mjpeg_decode_sos(s); - /* buggy avid puts EOI every 10-20th frame */ - /* if restart period is over process EOI */ - if ((s->buggy_avid && !s->interlaced) || s->restart_interval) - goto eoi_parser; + /* buggy avid puts EOI every 10-20th frame */ + /* if restart period is over process EOI */ + if ((s->buggy_avid && !s->interlaced) || s->restart_interval) + goto eoi_parser; + break; + case DRI: + mjpeg_decode_dri(s); break; - case DRI: - mjpeg_decode_dri(s); - break; - case SOF1: - case SOF2: - case SOF5: - case SOF6: - case SOF7: - case SOF9: - case SOF10: - case SOF11: - case SOF13: - case SOF14: - case SOF15: - case JPG: - av_log(s->avctx, AV_LOG_ERROR, "mjpeg: unsupported coding type (%x)\n", start_code); - break; -// default: -// printf("mjpeg: unsupported marker (%x)\n", start_code); -// break; + case SOF1: + case SOF2: + case SOF5: + case SOF6: + case SOF7: + case SOF9: + case SOF10: + case SOF11: + case SOF13: + case SOF14: + case SOF15: + case JPG: + av_log(s->avctx, AV_LOG_ERROR, "mjpeg: unsupported coding type (%x)\n", start_code); + break; +// default: +// printf("mjpeg: unsupported marker (%x)\n", start_code); +// break; } not_the_end: - /* eof process start code */ - buf_ptr += (get_bits_count(&s->gb)+7)/8; - dprintf("marker parser used %d bytes (%d bits)\n", - (get_bits_count(&s->gb)+7)/8, get_bits_count(&s->gb)); + /* eof process start code */ + buf_ptr += (get_bits_count(&s->gb)+7)/8; + dprintf("marker parser used %d bytes (%d bits)\n", + (get_bits_count(&s->gb)+7)/8, get_bits_count(&s->gb)); } } } @@ -2057,8 +2057,8 @@ if (get_bits_long(&hgb, 32) != be2me_32(ff_get_fourcc("mjpg"))) { - dprintf("not mjpeg-b (bad fourcc)\n"); - return 0; + dprintf("not mjpeg-b (bad fourcc)\n"); + return 0; } field_size = get_bits_long(&hgb, 32); /* field size */ @@ -2067,34 +2067,34 @@ second_field_offs = get_bits_long(&hgb, 32); dprintf("second field offs: 0x%x\n", second_field_offs); if (second_field_offs) - s->interlaced = 1; + s->interlaced = 1; dqt_offs = get_bits_long(&hgb, 32); dprintf("dqt offs: 0x%x\n", dqt_offs); if (dqt_offs) { - init_get_bits(&s->gb, buf+dqt_offs, (buf_end - (buf+dqt_offs))*8); - s->start_code = DQT; - mjpeg_decode_dqt(s); + init_get_bits(&s->gb, buf+dqt_offs, (buf_end - (buf+dqt_offs))*8); + s->start_code = DQT; + mjpeg_decode_dqt(s); } dht_offs = get_bits_long(&hgb, 32); dprintf("dht offs: 0x%x\n", dht_offs); if (dht_offs) { - init_get_bits(&s->gb, buf+dht_offs, (buf_end - (buf+dht_offs))*8); - s->start_code = DHT; - mjpeg_decode_dht(s); + init_get_bits(&s->gb, buf+dht_offs, (buf_end - (buf+dht_offs))*8); + s->start_code = DHT; + mjpeg_decode_dht(s); } sof_offs = get_bits_long(&hgb, 32); dprintf("sof offs: 0x%x\n", sof_offs); if (sof_offs) { - init_get_bits(&s->gb, buf+sof_offs, (buf_end - (buf+sof_offs))*8); - s->start_code = SOF0; - if (mjpeg_decode_sof(s) < 0) - return -1; + init_get_bits(&s->gb, buf+sof_offs, (buf_end - (buf+sof_offs))*8); + s->start_code = SOF0; + if (mjpeg_decode_sof(s) < 0) + return -1; } sos_offs = get_bits_long(&hgb, 32); @@ -2103,22 +2103,22 @@ dprintf("sod offs: 0x%x\n", sod_offs); if (sos_offs) { -// init_get_bits(&s->gb, buf+sos_offs, (buf_end - (buf+sos_offs))*8); - init_get_bits(&s->gb, buf+sos_offs, field_size*8); - s->mjpb_skiptosod = (sod_offs - sos_offs - show_bits(&s->gb, 16)); - s->start_code = SOS; - mjpeg_decode_sos(s); +// init_get_bits(&s->gb, buf+sos_offs, (buf_end - (buf+sos_offs))*8); + init_get_bits(&s->gb, buf+sos_offs, field_size*8); + s->mjpb_skiptosod = (sod_offs - sos_offs - show_bits(&s->gb, 16)); + s->start_code = SOS; + mjpeg_decode_sos(s); } if (s->interlaced) { s->bottom_field ^= 1; /* if not bottom field, do not output image yet */ if (s->bottom_field && second_field_offs) - { - buf_ptr = buf + second_field_offs; - second_field_offs = 0; - goto read_header; - } + { + buf_ptr = buf + second_field_offs; + second_field_offs = 0; + goto read_header; + } } //XXX FIXME factorize, this looks very similar to the EOI code @@ -2153,7 +2153,7 @@ int i = 0, j = 0; if (!avctx->width || !avctx->height) - return -1; + return -1; buf_ptr = buf; buf_end = buf + buf_size; @@ -2161,7 +2161,7 @@ #if 1 recoded = av_mallocz(buf_size + 1024); if (!recoded) - return -1; + return -1; /* SOI */ recoded[j++] = 0xFF; @@ -2187,9 +2187,9 @@ for (i = 14; i < buf_size && j < buf_size+1024-2; i++) { - recoded[j++] = buf[i]; - if (buf[i] == 0xff) - recoded[j++] = 0; + recoded[j++] = buf[i]; + if (buf[i] == 0xff) + recoded[j++] = 0; } /* EOI */ @@ -2229,33 +2229,33 @@ if (avctx->get_buffer(avctx, &s->picture) < 0) { av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); - return -1; + return -1; } s->picture.pict_type = I_TYPE; s->picture.key_frame = 1; for (i = 0; i < 3; i++) - s->linesize[i] = s->picture.linesize[i] << s->interlaced; + s->linesize[i] = s->picture.linesize[i] << s->interlaced; /* DQT */ for (i = 0; i < 64; i++) { - j = s->scantable.permutated[i]; - s->quant_matrixes[0][j] = sp5x_quant_table[(qscale * 2) + i]; + j = s->scantable.permutated[i]; + s->quant_matrixes[0][j] = sp5x_quant_table[(qscale * 2) + i]; } s->qscale[0] = FFMAX( - s->quant_matrixes[0][s->scantable.permutated[1]], - s->quant_matrixes[0][s->scantable.permutated[8]]) >> 1; + s->quant_matrixes[0][s->scantable.permutated[1]], + s->quant_matrixes[0][s->scantable.permutated[8]]) >> 1; for (i = 0; i < 64; i++) { - j = s->scantable.permutated[i]; - s->quant_matrixes[1][j] = sp5x_quant_table[(qscale * 2) + 1 + i]; + j = s->scantable.permutated[i]; + s->quant_matrixes[1][j] = sp5x_quant_table[(qscale * 2) + 1 + i]; } s->qscale[1] = FFMAX( - s->quant_matrixes[1][s->scantable.permutated[1]], - s->quant_matrixes[1][s->scantable.permutated[8]]) >> 1; + s->quant_matrixes[1][s->scantable.permutated[1]], + s->quant_matrixes[1][s->scantable.permutated[8]]) >> 1; /* DHT */ @@ -2282,7 +2282,7 @@ s->ac_index[2] = 1; for (i = 0; i < 3; i++) - s->last_dc[i] = 1024; + s->last_dc[i] = 1024; s->mb_width = (s->width * s->h_max * 8 -1) / (s->h_max * 8); s->mb_height = (s->height * s->v_max * 8 -1) / (s->v_max * 8);
--- a/mlib/dsputil_mlib.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mlib/dsputil_mlib.c Thu Dec 22 01:10:11 2005 +0000 @@ -61,7 +61,7 @@ /* put block, width 16 pixel, height 8/16 */ static void put_pixels16_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -78,7 +78,7 @@ } static void put_pixels16_x2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -95,7 +95,7 @@ } static void put_pixels16_y2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -112,7 +112,7 @@ } static void put_pixels16_xy2_mlib(uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -131,7 +131,7 @@ /* put block, width 8 pixel, height 4/8/16 */ static void put_pixels8_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -152,7 +152,7 @@ } static void put_pixels8_x2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -173,7 +173,7 @@ } static void put_pixels8_y2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -194,7 +194,7 @@ } static void put_pixels8_xy2_mlib(uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -217,7 +217,7 @@ /* average block, width 16 pixel, height 8/16 */ static void avg_pixels16_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -234,7 +234,7 @@ } static void avg_pixels16_x2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -251,7 +251,7 @@ } static void avg_pixels16_y2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -268,7 +268,7 @@ } static void avg_pixels16_xy2_mlib(uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 8: @@ -287,7 +287,7 @@ /* average block, width 8 pixel, height 4/8/16 */ static void avg_pixels8_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -308,7 +308,7 @@ } static void avg_pixels8_x2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -329,7 +329,7 @@ } static void avg_pixels8_y2_mlib (uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -350,7 +350,7 @@ } static void avg_pixels8_xy2_mlib(uint8_t * dest, const uint8_t * ref, - int stride, int height) + int stride, int height) { switch (height) { case 4: @@ -450,7 +450,7 @@ void MPV_common_init_mlib(MpegEncContext *s) { if(s->avctx->dct_algo==FF_DCT_AUTO || s->avctx->dct_algo==FF_DCT_MLIB){ - s->dsp.fdct = ff_fdct_mlib; + s->dsp.fdct = ff_fdct_mlib; } if(s->avctx->idct_algo==FF_IDCT_AUTO || s->avctx->idct_algo==FF_IDCT_MLIB){
--- a/motion_est.c Wed Dec 21 17:50:40 2005 +0000 +++ b/motion_est.c Thu Dec 22 01:10:11 2005 +0000 @@ -45,7 +45,7 @@ #define P_MV1 P[9] static inline int sad_hpel_motion_search(MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h); @@ -293,25 +293,25 @@ s = 0; for (i = 0; i < 16; i++) { - for (j = 0; j < 16; j += 8) { - s += ABS(pix[0]-mean); - s += ABS(pix[1]-mean); - s += ABS(pix[2]-mean); - s += ABS(pix[3]-mean); - s += ABS(pix[4]-mean); - s += ABS(pix[5]-mean); - s += ABS(pix[6]-mean); - s += ABS(pix[7]-mean); - pix += 8; - } - pix += line_size - 16; + for (j = 0; j < 16; j += 8) { + s += ABS(pix[0]-mean); + s += ABS(pix[1]-mean); + s += ABS(pix[2]-mean); + s += ABS(pix[3]-mean); + s += ABS(pix[4]-mean); + s += ABS(pix[5]-mean); + s += ABS(pix[6]-mean); + s += ABS(pix[7]-mean); + pix += 8; + } + pix += line_size - 16; } return s; } #endif static inline void no_motion_search(MpegEncContext * s, - int *mx_ptr, int *my_ptr) + int *mx_ptr, int *my_ptr) { *mx_ptr = 16 * s->mb_x; *my_ptr = 16 * s->mb_y; @@ -328,35 +328,35 @@ xx = 16 * s->mb_x; yy = 16 * s->mb_y; - x1 = xx - range + 1; /* we loose one pixel to avoid boundary pb with half pixel pred */ + x1 = xx - range + 1; /* we loose one pixel to avoid boundary pb with half pixel pred */ if (x1 < xmin) - x1 = xmin; + x1 = xmin; x2 = xx + range - 1; if (x2 > xmax) - x2 = xmax; + x2 = xmax; y1 = yy - range + 1; if (y1 < ymin) - y1 = ymin; + y1 = ymin; y2 = yy + range - 1; if (y2 > ymax) - y2 = ymax; + y2 = ymax; pix = s->new_picture.data[0] + (yy * s->linesize) + xx; dmin = 0x7fffffff; mx = 0; my = 0; for (y = y1; y <= y2; y++) { - for (x = x1; x <= x2; x++) { - d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, - s->linesize, 16); - if (d < dmin || - (d == dmin && - (abs(x - xx) + abs(y - yy)) < - (abs(mx - xx) + abs(my - yy)))) { - dmin = d; - mx = x; - my = y; - } - } + for (x = x1; x <= x2; x++) { + d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, + s->linesize, 16); + if (d < dmin || + (d == dmin && + (abs(x - xx) + abs(y - yy)) < + (abs(mx - xx) + abs(my - yy)))) { + dmin = d; + mx = x; + my = y; + } + } } *mx_ptr = mx; @@ -364,8 +364,8 @@ #if 0 if (*mx_ptr < -(2 * range) || *mx_ptr >= (2 * range) || - *my_ptr < -(2 * range) || *my_ptr >= (2 * range)) { - fprintf(stderr, "error %d %d\n", *mx_ptr, *my_ptr); + *my_ptr < -(2 * range) || *my_ptr >= (2 * range)) { + fprintf(stderr, "error %d %d\n", *mx_ptr, *my_ptr); } #endif return dmin; @@ -386,22 +386,22 @@ /* Left limit */ x1 = xx - range; if (x1 < xmin) - x1 = xmin; + x1 = xmin; /* Right limit */ x2 = xx + range; if (x2 > xmax) - x2 = xmax; + x2 = xmax; /* Upper limit */ y1 = yy - range; if (y1 < ymin) - y1 = ymin; + y1 = ymin; /* Lower limit */ y2 = yy + range; if (y2 > ymax) - y2 = ymax; + y2 = ymax; pix = s->new_picture.data[0] + (yy * s->linesize) + xx; dmin = 0x7fffffff; @@ -409,34 +409,34 @@ my = 0; do { - for (y = y1; y <= y2; y += range) { - for (x = x1; x <= x2; x += range) { - d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); - if (d < dmin || (d == dmin && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { - dmin = d; - mx = x; - my = y; - } - } - } + for (y = y1; y <= y2; y += range) { + for (x = x1; x <= x2; x += range) { + d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); + if (d < dmin || (d == dmin && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { + dmin = d; + mx = x; + my = y; + } + } + } - range = range >> 1; + range = range >> 1; - x1 = mx - range; - if (x1 < xmin) - x1 = xmin; + x1 = mx - range; + if (x1 < xmin) + x1 = xmin; - x2 = mx + range; - if (x2 > xmax) - x2 = xmax; + x2 = mx + range; + if (x2 > xmax) + x2 = xmax; - y1 = my - range; - if (y1 < ymin) - y1 = ymin; + y1 = my - range; + if (y1 < ymin) + y1 = ymin; - y2 = my + range; - if (y2 > ymax) - y2 = ymax; + y2 = my + range; + if (y2 > ymax) + y2 = ymax; } while (range >= 1); @@ -462,22 +462,22 @@ /* Left limit */ x1 = xx - range; if (x1 < xmin) - x1 = xmin; + x1 = xmin; /* Right limit */ x2 = xx + range; if (x2 > xmax) - x2 = xmax; + x2 = xmax; /* Upper limit */ y1 = yy - range; if (y1 < ymin) - y1 = ymin; + y1 = ymin; /* Lower limit */ y2 = yy + range; if (y2 > ymax) - y2 = ymax; + y2 = ymax; pix = s->new_picture.data[0] + (yy * s->linesize) + xx; mx = 0; @@ -489,43 +489,43 @@ dminx = 0x7fffffff; dminy = 0x7fffffff; - lastx = x; - for (x = x1; x <= x2; x += range) { - d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); - if (d < dminx || (d == dminx && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { - dminx = d; - mx = x; - } - } + lastx = x; + for (x = x1; x <= x2; x += range) { + d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); + if (d < dminx || (d == dminx && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { + dminx = d; + mx = x; + } + } - x = lastx; - for (y = y1; y <= y2; y += range) { - d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); - if (d < dminy || (d == dminy && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { - dminy = d; - my = y; - } - } + x = lastx; + for (y = y1; y <= y2; y += range) { + d = s->dsp.pix_abs[0][0](NULL, pix, ref_picture + (y * s->linesize) + x, s->linesize, 16); + if (d < dminy || (d == dminy && (abs(x - xx) + abs(y - yy)) < (abs(mx - xx) + abs(my - yy)))) { + dminy = d; + my = y; + } + } - range = range >> 1; + range = range >> 1; - x = mx; - y = my; - x1 = mx - range; - if (x1 < xmin) - x1 = xmin; + x = mx; + y = my; + x1 = mx - range; + if (x1 < xmin) + x1 = xmin; - x2 = mx + range; - if (x2 > xmax) - x2 = xmax; + x2 = mx + range; + if (x2 > xmax) + x2 = xmax; - y1 = my - range; - if (y1 < ymin) - y1 = ymin; + y1 = my - range; + if (y1 < ymin) + y1 = ymin; - y2 = my + range; - if (y2 > ymax) - y2 = ymax; + y2 = my + range; + if (y2 > ymax) + y2 = ymax; } while (range >= 1); @@ -550,7 +550,7 @@ } static inline int sad_hpel_motion_search(MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h) { @@ -1190,24 +1190,24 @@ switch(s->me_method) { case ME_ZERO: default: - no_motion_search(s, &mx, &my); + no_motion_search(s, &mx, &my); mx-= mb_x*16; my-= mb_y*16; dmin = 0; break; #if 0 case ME_FULL: - dmin = full_motion_search(s, &mx, &my, range, ref_picture); + dmin = full_motion_search(s, &mx, &my, range, ref_picture); mx-= mb_x*16; my-= mb_y*16; break; case ME_LOG: - dmin = log_motion_search(s, &mx, &my, range / 2, ref_picture); + dmin = log_motion_search(s, &mx, &my, range / 2, ref_picture); mx-= mb_x*16; my-= mb_y*16; break; case ME_PHODS: - dmin = phods_motion_search(s, &mx, &my, range / 2, ref_picture); + dmin = phods_motion_search(s, &mx, &my, range / 2, ref_picture); mx-= mb_x*16; my-= mb_y*16; break; @@ -1264,7 +1264,7 @@ #if 0 printf("varc=%4d avg_var=%4d (sum=%4d) vard=%4d mx=%2d my=%2d\n", - varc, s->avg_mb_var, sum, vard, mx - xx, my - yy); + varc, s->avg_mb_var, sum, vard, mx - xx, my - yy); #endif if(mb_type){ if (vard <= 64 || vard < varc) @@ -1479,24 +1479,24 @@ switch(s->me_method) { case ME_ZERO: default: - no_motion_search(s, &mx, &my); + no_motion_search(s, &mx, &my); dmin = 0; mx-= mb_x*16; my-= mb_y*16; break; #if 0 case ME_FULL: - dmin = full_motion_search(s, &mx, &my, range, ref_picture); + dmin = full_motion_search(s, &mx, &my, range, ref_picture); mx-= mb_x*16; my-= mb_y*16; break; case ME_LOG: - dmin = log_motion_search(s, &mx, &my, range / 2, ref_picture); + dmin = log_motion_search(s, &mx, &my, range / 2, ref_picture); mx-= mb_x*16; my-= mb_y*16; break; case ME_PHODS: - dmin = phods_motion_search(s, &mx, &my, range / 2, ref_picture); + dmin = phods_motion_search(s, &mx, &my, range / 2, ref_picture); mx-= mb_x*16; my-= mb_y*16; break;
--- a/motion_est_template.c Wed Dec 21 17:50:40 2005 +0000 +++ b/motion_est_template.c Thu Dec 22 01:10:11 2005 +0000 @@ -45,7 +45,7 @@ #if 0 static int hpel_motion_search)(MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, uint8_t *ref_data[3], int size) { @@ -113,7 +113,7 @@ #else static int hpel_motion_search(MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h) { @@ -271,7 +271,7 @@ } static int qpel_motion_search(MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h) { @@ -1005,7 +1005,7 @@ //printf("%d %d %d %d //",xmin, ymin, xmax, ymax); /* first line */ if (s->first_slice_line) { - CHECK_MV(P_LEFT[0]>>shift, P_LEFT[1]>>shift) + CHECK_MV(P_LEFT[0]>>shift, P_LEFT[1]>>shift) CHECK_CLIPED_MV((last_mv[ref_mv_xy][0]*ref_mv_scale + (1<<15))>>16, (last_mv[ref_mv_xy][1]*ref_mv_scale + (1<<15))>>16) CHECK_MV(P_MV1[0]>>shift, P_MV1[1]>>shift) @@ -1067,7 +1067,7 @@ //printf("%d %d %d %d //",xmin, ymin, xmax, ymax); /* first line */ if (s->first_slice_line) { - CHECK_MV(P_LEFT[0]>>shift, P_LEFT[1]>>shift) + CHECK_MV(P_LEFT[0]>>shift, P_LEFT[1]>>shift) CHECK_CLIPED_MV((last_mv[ref_mv_xy][0]*ref_mv_scale + (1<<15))>>16, (last_mv[ref_mv_xy][1]*ref_mv_scale + (1<<15))>>16) CHECK_MV(P_MV1[0]>>shift, P_MV1[1]>>shift)
--- a/mp3lameaudio.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mp3lameaudio.c Thu Dec 22 01:10:11 2005 +0000 @@ -28,51 +28,51 @@ #define BUFFER_SIZE (2*MPA_FRAME_SIZE) typedef struct Mp3AudioContext { - lame_global_flags *gfp; - int stereo; + lame_global_flags *gfp; + int stereo; uint8_t buffer[BUFFER_SIZE]; int buffer_index; } Mp3AudioContext; static int MP3lame_encode_init(AVCodecContext *avctx) { - Mp3AudioContext *s = avctx->priv_data; + Mp3AudioContext *s = avctx->priv_data; - if (avctx->channels > 2) - return -1; + if (avctx->channels > 2) + return -1; - s->stereo = avctx->channels > 1 ? 1 : 0; + s->stereo = avctx->channels > 1 ? 1 : 0; - if ((s->gfp = lame_init()) == NULL) - goto err; - lame_set_in_samplerate(s->gfp, avctx->sample_rate); - lame_set_out_samplerate(s->gfp, avctx->sample_rate); - lame_set_num_channels(s->gfp, avctx->channels); - /* lame 3.91 dies on quality != 5 */ - lame_set_quality(s->gfp, 5); - /* lame 3.91 doesn't work in mono */ - lame_set_mode(s->gfp, JOINT_STEREO); - lame_set_brate(s->gfp, avctx->bit_rate/1000); + if ((s->gfp = lame_init()) == NULL) + goto err; + lame_set_in_samplerate(s->gfp, avctx->sample_rate); + lame_set_out_samplerate(s->gfp, avctx->sample_rate); + lame_set_num_channels(s->gfp, avctx->channels); + /* lame 3.91 dies on quality != 5 */ + lame_set_quality(s->gfp, 5); + /* lame 3.91 doesn't work in mono */ + lame_set_mode(s->gfp, JOINT_STEREO); + lame_set_brate(s->gfp, avctx->bit_rate/1000); if(avctx->flags & CODEC_FLAG_QSCALE) { lame_set_brate(s->gfp, 0); lame_set_VBR(s->gfp, vbr_default); lame_set_VBR_q(s->gfp, avctx->global_quality / (float)FF_QP2LAMBDA); } lame_set_bWriteVbrTag(s->gfp,0); - if (lame_init_params(s->gfp) < 0) - goto err_close; + if (lame_init_params(s->gfp) < 0) + goto err_close; - avctx->frame_size = lame_get_framesize(s->gfp); + avctx->frame_size = lame_get_framesize(s->gfp); avctx->coded_frame= avcodec_alloc_frame(); avctx->coded_frame->key_frame= 1; - return 0; + return 0; err_close: - lame_close(s->gfp); + lame_close(s->gfp); err: - return -1; + return -1; } static const int sSampleRates[3] = { @@ -136,11 +136,11 @@ int MP3lame_encode_frame(AVCodecContext *avctx, unsigned char *frame, int buf_size, void *data) { - Mp3AudioContext *s = avctx->priv_data; - int len; - int lame_result; + Mp3AudioContext *s = avctx->priv_data; + int len; + int lame_result; - /* lame 3.91 dies on '1-channel interleaved' data */ + /* lame 3.91 dies on '1-channel interleaved' data */ if(data){ if (s->stereo) { @@ -198,12 +198,12 @@ int MP3lame_encode_close(AVCodecContext *avctx) { - Mp3AudioContext *s = avctx->priv_data; + Mp3AudioContext *s = avctx->priv_data; av_freep(&avctx->coded_frame); - lame_close(s->gfp); - return 0; + lame_close(s->gfp); + return 0; }
--- a/mpeg12.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mpeg12.c Thu Dec 22 01:10:11 2005 +0000 @@ -35,14 +35,14 @@ /* Start codes. */ -#define SEQ_END_CODE 0x000001b7 -#define SEQ_START_CODE 0x000001b3 -#define GOP_START_CODE 0x000001b8 -#define PICTURE_START_CODE 0x00000100 -#define SLICE_MIN_START_CODE 0x00000101 -#define SLICE_MAX_START_CODE 0x000001af -#define EXT_START_CODE 0x000001b5 -#define USER_START_CODE 0x000001b2 +#define SEQ_END_CODE 0x000001b7 +#define SEQ_START_CODE 0x000001b3 +#define GOP_START_CODE 0x000001b8 +#define PICTURE_START_CODE 0x00000100 +#define SLICE_MIN_START_CODE 0x00000101 +#define SLICE_MAX_START_CODE 0x000001af +#define EXT_START_CODE 0x000001b5 +#define USER_START_CODE 0x000001b2 #define DC_VLC_BITS 9 #define MV_VLC_BITS 9 @@ -89,7 +89,7 @@ const enum PixelFormat pixfmt_xvmc_mpg2_420[] = { PIX_FMT_XVMC_MPEG2_IDCT, PIX_FMT_XVMC_MPEG2_MC, - -1}; + -1}; #ifdef CONFIG_ENCODERS static uint8_t (*mv_penalty)[MAX_MV*2+1]= NULL; static uint8_t fcode_tab[MAX_MV*2+1]; @@ -166,7 +166,7 @@ code= rl->index_run[0][run] + alevel - 1; if (code < 111 /* rl->n */) { - /* store the vlc & sign at once */ + /* store the vlc & sign at once */ len= mpeg1_vlc[code][1]+1; bits= (mpeg1_vlc[code][0]<<1) + sign; } else { @@ -764,38 +764,38 @@ if(!done){ int f_code; int mv; - int i; + int i; done=1; init_rl(&rl_mpeg1, 1); - for(i=0; i<64; i++) - { - mpeg1_max_level[0][i]= rl_mpeg1.max_level[0][i]; - mpeg1_index_run[0][i]= rl_mpeg1.index_run[0][i]; - } + for(i=0; i<64; i++) + { + mpeg1_max_level[0][i]= rl_mpeg1.max_level[0][i]; + mpeg1_index_run[0][i]= rl_mpeg1.index_run[0][i]; + } init_uni_ac_vlc(&rl_mpeg1, uni_mpeg1_ac_vlc_bits, uni_mpeg1_ac_vlc_len); - /* build unified dc encoding tables */ - for(i=-255; i<256; i++) - { - int adiff, index; - int bits, code; - int diff=i; - - adiff = ABS(diff); - if(diff<0) diff--; - index = av_log2(2*adiff); - - bits= vlc_dc_lum_bits[index] + index; - code= (vlc_dc_lum_code[index]<<index) + (diff & ((1 << index) - 1)); - mpeg1_lum_dc_uni[i+255]= bits + (code<<8); - - bits= vlc_dc_chroma_bits[index] + index; - code= (vlc_dc_chroma_code[index]<<index) + (diff & ((1 << index) - 1)); - mpeg1_chr_dc_uni[i+255]= bits + (code<<8); - } + /* build unified dc encoding tables */ + for(i=-255; i<256; i++) + { + int adiff, index; + int bits, code; + int diff=i; + + adiff = ABS(diff); + if(diff<0) diff--; + index = av_log2(2*adiff); + + bits= vlc_dc_lum_bits[index] + index; + code= (vlc_dc_lum_code[index]<<index) + (diff & ((1 << index) - 1)); + mpeg1_lum_dc_uni[i+255]= bits + (code<<8); + + bits= vlc_dc_chroma_bits[index] + index; + code= (vlc_dc_chroma_code[index]<<index) + (diff & ((1 << index) - 1)); + mpeg1_chr_dc_uni[i+255]= bits + (code<<8); + } mv_penalty= av_mallocz( sizeof(uint8_t)*(MAX_FCODE+1)*(2*MAX_MV+1) ); @@ -873,14 +873,14 @@ }else{ if (component == 0) { put_bits( - &s->pb, - mpeg1_lum_dc_uni[diff+255]&0xFF, - mpeg1_lum_dc_uni[diff+255]>>8); + &s->pb, + mpeg1_lum_dc_uni[diff+255]&0xFF, + mpeg1_lum_dc_uni[diff+255]>>8); } else { put_bits( &s->pb, - mpeg1_chr_dc_uni[diff+255]&0xFF, - mpeg1_chr_dc_uni[diff+255]>>8); + mpeg1_chr_dc_uni[diff+255]&0xFF, + mpeg1_chr_dc_uni[diff+255]>>8); } } } @@ -946,10 +946,10 @@ // code = get_rl_index(rl, 0, run, alevel); if (alevel <= mpeg1_max_level[0][run]){ code= mpeg1_index_run[0][run] + alevel - 1; - /* store the vlc & sign at once */ + /* store the vlc & sign at once */ put_bits(&s->pb, mpeg1_vlc[code][1]+1, (mpeg1_vlc[code][0]<<1) + sign); } else { - /* escape seems to be pretty rare <5% so i dont optimize it */ + /* escape seems to be pretty rare <5% so i dont optimize it */ put_bits(&s->pb, mpeg1_vlc[111/*rl->n*/][1], mpeg1_vlc[111/*rl->n*/][0]); /* escape: only clip in this case */ put_bits(&s->pb, 6, run); @@ -1376,8 +1376,8 @@ return -1; } if(mb_block_count > 6){ - cbp<<= mb_block_count-6; - cbp |= get_bits(&s->gb, mb_block_count-6); + cbp<<= mb_block_count-6; + cbp |= get_bits(&s->gb, mb_block_count-6); } #ifdef HAVE_XVMC @@ -2074,7 +2074,7 @@ uint8_t old_permutation[64]; if ( - (s1->mpeg_enc_ctx_allocated == 0)|| + (s1->mpeg_enc_ctx_allocated == 0)|| avctx->coded_width != s->width || avctx->coded_height != s->height|| s1->save_aspect_info != s->aspect_ratio_info|| @@ -2088,8 +2088,8 @@ s->parse_context= pc; } - if( (s->width == 0 )||(s->height == 0)) - return -2; + if( (s->width == 0 )||(s->height == 0)) + return -2; avcodec_set_dimensions(avctx, s->width, s->height); avctx->bit_rate = s->bit_rate; @@ -2129,7 +2129,7 @@ mpeg2_aspect[s->aspect_ratio_info], (AVRational){s1->pan_scan.width, s1->pan_scan.height} ); - } + } }else{ s->avctx->sample_aspect_ratio= mpeg2_aspect[s->aspect_ratio_info]; @@ -2312,16 +2312,16 @@ nofco = 1; if(s->progressive_sequence){ if(s->repeat_first_field){ - nofco++; - if(s->top_field_first) - nofco++; - } + nofco++; + if(s->top_field_first) + nofco++; + } }else{ if(s->picture_structure == PICT_FRAME){ nofco++; - if(s->repeat_first_field) - nofco++; - } + if(s->repeat_first_field) + nofco++; + } } for(i=0; i<nofco; i++){ s1->pan_scan.position[i][0]= get_sbits(&s->gb, 16); @@ -2985,8 +2985,8 @@ if(s->avctx->debug & FF_DEBUG_PICT_INFO) av_log(s->avctx, AV_LOG_DEBUG, "GOP (%2d:%02d:%02d.[%02d]) broken_link=%d\n", - time_code_hours, time_code_minutes, time_code_seconds, - time_code_pictures, broken_link); + time_code_hours, time_code_minutes, time_code_seconds, + time_code_pictures, broken_link); } /** * finds the end of the current frame in the bitstream. @@ -3044,13 +3044,13 @@ dprintf("fill_buffer\n"); if (buf_size == 0) { - /* special case for last picture */ - if (s2->low_delay==0 && s2->next_picture_ptr) { - *picture= *(AVFrame*)s2->next_picture_ptr; - s2->next_picture_ptr= NULL; - - *data_size = sizeof(AVFrame); - } + /* special case for last picture */ + if (s2->low_delay==0 && s2->next_picture_ptr) { + *picture= *(AVFrame*)s2->next_picture_ptr; + s2->next_picture_ptr= NULL; + + *data_size = sizeof(AVFrame); + } return 0; } @@ -3111,13 +3111,13 @@ switch(start_code) { case SEQ_START_CODE: mpeg1_decode_sequence(avctx, buf_ptr, - input_size); + input_size); break; case PICTURE_START_CODE: /* we have a complete image : we try to decompress it */ mpeg1_decode_picture(avctx, - buf_ptr, input_size); + buf_ptr, input_size); break; case EXT_START_CODE: mpeg_decode_extension(avctx,
--- a/mpeg12data.h Wed Dec 21 17:50:40 2005 +0000 +++ b/mpeg12data.h Thu Dec 22 01:10:11 2005 +0000 @@ -4,14 +4,14 @@ */ const int16_t ff_mpeg1_default_intra_matrix[64] = { - 8, 16, 19, 22, 26, 27, 29, 34, - 16, 16, 22, 24, 27, 29, 34, 37, - 19, 22, 26, 27, 29, 34, 34, 38, - 22, 22, 26, 27, 29, 34, 37, 40, - 22, 26, 27, 29, 32, 35, 40, 48, - 26, 27, 29, 32, 35, 40, 48, 58, - 26, 27, 29, 34, 38, 46, 56, 69, - 27, 29, 35, 38, 46, 56, 69, 83 + 8, 16, 19, 22, 26, 27, 29, 34, + 16, 16, 22, 24, 27, 29, 34, 37, + 19, 22, 26, 27, 29, 34, 34, 38, + 22, 22, 26, 27, 29, 34, 37, 40, + 22, 26, 27, 29, 32, 35, 40, 48, + 26, 27, 29, 32, 35, 40, 48, 58, + 26, 27, 29, 34, 38, 46, 56, 69, + 27, 29, 35, 38, 46, 56, 69, 83 }; const int16_t ff_mpeg1_default_non_intra_matrix[64] = {
--- a/mpegaudio.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mpegaudio.c Thu Dec 22 01:10:11 2005 +0000 @@ -748,7 +748,7 @@ } static int MPA_encode_frame(AVCodecContext *avctx, - unsigned char *frame, int buf_size, void *data) + unsigned char *frame, int buf_size, void *data) { MpegAudioContext *s = avctx->priv_data; short *samples = data;
--- a/mpegaudio.h Wed Dec 21 17:50:40 2005 +0000 +++ b/mpegaudio.h Thu Dec 22 01:10:11 2005 +0000 @@ -55,7 +55,7 @@ int mpa_decode_header(AVCodecContext *avctx, uint32_t head); void ff_mpa_synth_init(MPA_INT *window); void ff_mpa_synth_filter(MPA_INT *synth_buf_ptr, int *synth_buf_offset, - MPA_INT *window, int *dither_state, + MPA_INT *window, int *dither_state, OUT_INT *samples, int incr, int32_t sb_samples[SBLIMIT]);
--- a/mpegaudiodec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mpegaudiodec.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,7 +64,7 @@ struct GranuleDef; typedef struct MPADecodeContext { - uint8_t inbuf1[2][MPA_MAX_CODED_FRAME_SIZE + BACKSTEP_SIZE]; /* input buffer */ + uint8_t inbuf1[2][MPA_MAX_CODED_FRAME_SIZE + BACKSTEP_SIZE]; /* input buffer */ int inbuf_index; uint8_t *inbuf_ptr, *inbuf; int frame_size; @@ -340,13 +340,13 @@ scale_factor_mult[i][2]); } - ff_mpa_synth_init(window); + ff_mpa_synth_init(window); /* huffman decode tables */ huff_code_table[0] = NULL; for(i=1;i<16;i++) { const HuffTable *h = &mpa_huff_tables[i]; - int xsize, x, y; + int xsize, x, y; unsigned int n; uint8_t *code_table; @@ -378,11 +378,11 @@ band_index_long[i][22] = k; } - /* compute n ^ (4/3) and store it in mantissa/exp format */ - table_4_3_exp= av_mallocz_static(TABLE_4_3_SIZE * sizeof(table_4_3_exp[0])); + /* compute n ^ (4/3) and store it in mantissa/exp format */ + table_4_3_exp= av_mallocz_static(TABLE_4_3_SIZE * sizeof(table_4_3_exp[0])); if(!table_4_3_exp) - return -1; - table_4_3_value= av_mallocz_static(TABLE_4_3_SIZE * sizeof(table_4_3_value[0])); + return -1; + table_4_3_value= av_mallocz_static(TABLE_4_3_SIZE * sizeof(table_4_3_value[0])); if(!table_4_3_value) return -1; @@ -844,7 +844,7 @@ 32 samples. */ /* XXX: optimize by avoiding ring buffer usage */ void ff_mpa_synth_filter(MPA_INT *synth_buf_ptr, int *synth_buf_offset, - MPA_INT *window, int *dither_state, + MPA_INT *window, int *dither_state, OUT_INT *samples, int incr, int32_t sb_samples[SBLIMIT]) { @@ -2440,8 +2440,8 @@ samples_ptr = samples + ch; for(i=0;i<nb_frames;i++) { ff_mpa_synth_filter(s->synth_buf[ch], &(s->synth_buf_offset[ch]), - window, &s->dither_state, - samples_ptr, s->nb_channels, + window, &s->dither_state, + samples_ptr, s->nb_channels, s->sb_samples[ch][i]); samples_ptr += 32 * s->nb_channels; } @@ -2453,8 +2453,8 @@ } static int decode_frame(AVCodecContext * avctx, - void *data, int *data_size, - uint8_t * buf, int buf_size) + void *data, int *data_size, + uint8_t * buf, int buf_size) { MPADecodeContext *s = avctx->priv_data; uint32_t header; @@ -2464,8 +2464,8 @@ buf_ptr = buf; while (buf_size > 0) { - len = s->inbuf_ptr - s->inbuf; - if (s->frame_size == 0) { + len = s->inbuf_ptr - s->inbuf; + if (s->frame_size == 0) { /* special case for next header for first frame in free format case (XXX: find a simpler method) */ if (s->free_format_next_header != 0) { @@ -2477,34 +2477,34 @@ s->free_format_next_header = 0; goto got_header; } - /* no header seen : find one. We need at least HEADER_SIZE + /* no header seen : find one. We need at least HEADER_SIZE bytes to parse it */ - len = HEADER_SIZE - len; - if (len > buf_size) - len = buf_size; - if (len > 0) { - memcpy(s->inbuf_ptr, buf_ptr, len); - buf_ptr += len; - buf_size -= len; - s->inbuf_ptr += len; - } - if ((s->inbuf_ptr - s->inbuf) >= HEADER_SIZE) { + len = HEADER_SIZE - len; + if (len > buf_size) + len = buf_size; + if (len > 0) { + memcpy(s->inbuf_ptr, buf_ptr, len); + buf_ptr += len; + buf_size -= len; + s->inbuf_ptr += len; + } + if ((s->inbuf_ptr - s->inbuf) >= HEADER_SIZE) { got_header: - header = (s->inbuf[0] << 24) | (s->inbuf[1] << 16) | - (s->inbuf[2] << 8) | s->inbuf[3]; + header = (s->inbuf[0] << 24) | (s->inbuf[1] << 16) | + (s->inbuf[2] << 8) | s->inbuf[3]; - if (ff_mpa_check_header(header) < 0) { - /* no sync found : move by one byte (inefficient, but simple!) */ - memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); - s->inbuf_ptr--; + if (ff_mpa_check_header(header) < 0) { + /* no sync found : move by one byte (inefficient, but simple!) */ + memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); + s->inbuf_ptr--; dprintf("skip %x\n", header); /* reset free format frame size to give a chance to get a new bitrate */ s->free_format_frame_size = 0; - } else { - if (decode_header(s, header) == 1) { + } else { + if (decode_header(s, header) == 1) { /* free format: prepare to compute frame size */ - s->frame_size = -1; + s->frame_size = -1; } /* update codec info */ avctx->sample_rate = s->sample_rate; @@ -2525,18 +2525,18 @@ avctx->frame_size = 1152; break; } - } - } + } + } } else if (s->frame_size == -1) { /* free format : find next sync to compute frame size */ - len = MPA_MAX_CODED_FRAME_SIZE - len; - if (len > buf_size) - len = buf_size; + len = MPA_MAX_CODED_FRAME_SIZE - len; + if (len > buf_size) + len = buf_size; if (len == 0) { - /* frame too long: resync */ + /* frame too long: resync */ s->frame_size = 0; - memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); - s->inbuf_ptr--; + memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); + s->inbuf_ptr--; } else { uint8_t *p, *pend; uint32_t header1; @@ -2580,17 +2580,17 @@ s->inbuf_ptr += len; buf_size -= len; } - } else if (len < s->frame_size) { + } else if (len < s->frame_size) { if (s->frame_size > MPA_MAX_CODED_FRAME_SIZE) s->frame_size = MPA_MAX_CODED_FRAME_SIZE; - len = s->frame_size - len; - if (len > buf_size) - len = buf_size; - memcpy(s->inbuf_ptr, buf_ptr, len); - buf_ptr += len; - s->inbuf_ptr += len; - buf_size -= len; - } + len = s->frame_size - len; + if (len > buf_size) + len = buf_size; + memcpy(s->inbuf_ptr, buf_ptr, len); + buf_ptr += len; + s->inbuf_ptr += len; + buf_size -= len; + } next_data: if (s->frame_size > 0 && (s->inbuf_ptr - s->inbuf) >= s->frame_size) { @@ -2601,22 +2601,22 @@ } else { out_size = mp_decode_frame(s, out_samples); } - s->inbuf_ptr = s->inbuf; - s->frame_size = 0; + s->inbuf_ptr = s->inbuf; + s->frame_size = 0; if(out_size>=0) - *data_size = out_size; + *data_size = out_size; else av_log(avctx, AV_LOG_DEBUG, "Error while decoding mpeg audio frame\n"); //FIXME return -1 / but also return the number of bytes consumed - break; - } + break; + } } return buf_ptr - buf; } static int decode_frame_adu(AVCodecContext * avctx, - void *data, int *data_size, - uint8_t * buf, int buf_size) + void *data, int *data_size, + uint8_t * buf, int buf_size) { MPADecodeContext *s = avctx->priv_data; uint32_t header; @@ -2747,8 +2747,8 @@ static int decode_frame_mp3on4(AVCodecContext * avctx, - void *data, int *data_size, - uint8_t * buf, int buf_size) + void *data, int *data_size, + uint8_t * buf, int buf_size) { MP3On4DecodeContext *s = avctx->priv_data; MPADecodeContext *m;
--- a/mpegvideo.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mpegvideo.c Thu Dec 22 01:10:11 2005 +0000 @@ -354,7 +354,7 @@ r= s->avctx->get_buffer(s->avctx, (AVFrame*)pic); if(r<0 || !pic->age || !pic->type || !pic->data[0]){ - av_log(s->avctx, AV_LOG_ERROR, "get_buffer() failed (%d %d %d %p)\n", r, pic->age, pic->type, pic->data[0]); + av_log(s->avctx, AV_LOG_ERROR, "get_buffer() failed (%d %d %d %p)\n", r, pic->age, pic->type, pic->data[0]); return -1; } @@ -913,7 +913,7 @@ s->width = avctx->width; s->height = avctx->height; if(avctx->gop_size > 600){ - av_log(avctx, AV_LOG_ERROR, "Warning keyframe interval too large! reducing it ...\n"); + av_log(avctx, AV_LOG_ERROR, "Warning keyframe interval too large! reducing it ...\n"); avctx->gop_size=600; } s->gop_size = avctx->gop_size; @@ -1120,7 +1120,7 @@ s->out_format = FMT_MJPEG; s->intra_only = 1; /* force intra only for jpeg */ s->mjpeg_write_tables = avctx->codec->id != CODEC_ID_JPEGLS; - s->mjpeg_data_only_frames = 0; /* write all the needed headers */ + s->mjpeg_data_only_frames = 0; /* write all the needed headers */ s->mjpeg_vsample[0] = 1<<chroma_v_shift; s->mjpeg_vsample[1] = 1; s->mjpeg_vsample[2] = 1; @@ -1143,24 +1143,24 @@ return -1; } s->out_format = FMT_H263; - s->obmc= (avctx->flags & CODEC_FLAG_OBMC) ? 1:0; + s->obmc= (avctx->flags & CODEC_FLAG_OBMC) ? 1:0; avctx->delay=0; s->low_delay=1; break; case CODEC_ID_H263P: s->out_format = FMT_H263; s->h263_plus = 1; - /* Fx */ + /* Fx */ s->umvplus = (avctx->flags & CODEC_FLAG_H263P_UMV) ? 1:0; - s->h263_aic= (avctx->flags & CODEC_FLAG_H263P_AIC) ? 1:0; - s->modified_quant= s->h263_aic; - s->alt_inter_vlc= (avctx->flags & CODEC_FLAG_H263P_AIV) ? 1:0; - s->obmc= (avctx->flags & CODEC_FLAG_OBMC) ? 1:0; - s->loop_filter= (avctx->flags & CODEC_FLAG_LOOP_FILTER) ? 1:0; - s->unrestricted_mv= s->obmc || s->loop_filter || s->umvplus; + s->h263_aic= (avctx->flags & CODEC_FLAG_H263P_AIC) ? 1:0; + s->modified_quant= s->h263_aic; + s->alt_inter_vlc= (avctx->flags & CODEC_FLAG_H263P_AIV) ? 1:0; + s->obmc= (avctx->flags & CODEC_FLAG_OBMC) ? 1:0; + s->loop_filter= (avctx->flags & CODEC_FLAG_LOOP_FILTER) ? 1:0; + s->unrestricted_mv= s->obmc || s->loop_filter || s->umvplus; s->h263_slice_structured= (s->flags & CODEC_FLAG_H263P_SLICE_STRUCT) ? 1:0; - /* /Fx */ + /* /Fx */ /* These are just to be sure */ avctx->delay=0; s->low_delay=1; @@ -2473,7 +2473,7 @@ dxy= ((motion_x>>3)&1) | ((motion_y>>2)&2); if (s->no_rounding){ - s->dsp.put_no_rnd_pixels_tab[0][dxy](dest_y, ptr, linesize, 16); + s->dsp.put_no_rnd_pixels_tab[0][dxy](dest_y, ptr, linesize, 16); }else{ s->dsp.put_pixels_tab [0][dxy](dest_y, ptr, linesize, 16); } @@ -4148,7 +4148,7 @@ } } - s->dsp.get_pixels(s->block[0], ptr_y , wrap_y); + s->dsp.get_pixels(s->block[0], ptr_y , wrap_y); s->dsp.get_pixels(s->block[1], ptr_y + 8, wrap_y); s->dsp.get_pixels(s->block[2], ptr_y + dct_offset , wrap_y); s->dsp.get_pixels(s->block[3], ptr_y + dct_offset + 8, wrap_y); @@ -4157,7 +4157,7 @@ skip_dct[4]= 1; skip_dct[5]= 1; }else{ - s->dsp.get_pixels(s->block[4], ptr_cb, wrap_c); + s->dsp.get_pixels(s->block[4], ptr_cb, wrap_c); s->dsp.get_pixels(s->block[5], ptr_cr, wrap_c); } }else{ @@ -4170,7 +4170,7 @@ dest_cr = s->dest[2]; if ((!s->no_rounding) || s->pict_type==B_TYPE){ - op_pix = s->dsp.put_pixels_tab; + op_pix = s->dsp.put_pixels_tab; op_qpix= s->dsp.put_qpel_pixels_tab; }else{ op_pix = s->dsp.put_no_rnd_pixels_tab; @@ -4208,7 +4208,7 @@ } } - s->dsp.diff_pixels(s->block[0], ptr_y , dest_y , wrap_y); + s->dsp.diff_pixels(s->block[0], ptr_y , dest_y , wrap_y); s->dsp.diff_pixels(s->block[1], ptr_y + 8, dest_y + 8, wrap_y); s->dsp.diff_pixels(s->block[2], ptr_y + dct_offset , dest_y + dct_offset , wrap_y); s->dsp.diff_pixels(s->block[3], ptr_y + dct_offset + 8, dest_y + dct_offset + 8, wrap_y); @@ -4223,7 +4223,7 @@ /* pre quantization */ if(s->current_picture.mc_mb_var[s->mb_stride*mb_y+ mb_x]<2*s->qscale*s->qscale){ //FIXME optimize - if(s->dsp.sad[1](NULL, ptr_y , dest_y , wrap_y, 8) < 20*s->qscale) skip_dct[0]= 1; + if(s->dsp.sad[1](NULL, ptr_y , dest_y , wrap_y, 8) < 20*s->qscale) skip_dct[0]= 1; if(s->dsp.sad[1](NULL, ptr_y + 8, dest_y + 8, wrap_y, 8) < 20*s->qscale) skip_dct[1]= 1; if(s->dsp.sad[1](NULL, ptr_y +dct_offset , dest_y +dct_offset , wrap_y, 8) < 20*s->qscale) skip_dct[2]= 1; if(s->dsp.sad[1](NULL, ptr_y +dct_offset+ 8, dest_y +dct_offset+ 8, wrap_y, 8) < 20*s->qscale) skip_dct[3]= 1; @@ -6265,7 +6265,7 @@ /* we need this permutation so that we correct the IDCT, we only permute the !=0 elements */ if (s->dsp.idct_permutation_type != FF_NO_IDCT_PERM) - ff_block_permute(block, s->dsp.idct_permutation, scantable, last_non_zero); + ff_block_permute(block, s->dsp.idct_permutation, scantable, last_non_zero); return last_non_zero; }
--- a/mpegvideo.h Wed Dec 21 17:50:40 2005 +0000 +++ b/mpegvideo.h Thu Dec 22 01:10:11 2005 +0000 @@ -126,7 +126,7 @@ uint8_t permutated[64]; uint8_t raster_end[64]; #ifdef ARCH_POWERPC - /** Used by dct_quantise_alitvec to find last-non-zero */ + /** Used by dct_quantise_alitvec to find last-non-zero */ uint8_t __align8 inverse[64]; #endif } ScanTable; @@ -181,7 +181,7 @@ uint16_t *mb_var; ///< Table for MB variances uint16_t *mc_mb_var; ///< Table for motion compensated MB variances uint8_t *mb_mean; ///< Table for MB luminance - int32_t *mb_cmp_score; ///< Table for MB cmp scores, for mb decision FIXME remove + int32_t *mb_cmp_score; ///< Table for MB cmp scores, for mb decision FIXME remove int b_frame_score; /* */ } Picture; @@ -245,7 +245,7 @@ uint8_t (*mv_penalty)[MAX_MV*2+1]; ///< amount of bits needed to encode a MV uint8_t *current_mv_penalty; int (*sub_motion_search)(struct MpegEncContext * s, - int *mx_ptr, int *my_ptr, int dmin, + int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h); }MotionEstContext;
--- a/msmpeg4.c Wed Dec 21 17:50:40 2005 +0000 +++ b/msmpeg4.c Thu Dec 22 01:10:11 2005 +0000 @@ -544,24 +544,24 @@ handle_slices(s); if (!s->mb_intra) { - /* compute cbp */ + /* compute cbp */ set_stat(ST_INTER_MB); - cbp = 0; - for (i = 0; i < 6; i++) { - if (s->block_last_index[i] >= 0) - cbp |= 1 << (5 - i); - } - if (s->use_skip_mb_code && (cbp | motion_x | motion_y) == 0) { - /* skip macroblock */ - put_bits(&s->pb, 1, 1); + cbp = 0; + for (i = 0; i < 6; i++) { + if (s->block_last_index[i] >= 0) + cbp |= 1 << (5 - i); + } + if (s->use_skip_mb_code && (cbp | motion_x | motion_y) == 0) { + /* skip macroblock */ + put_bits(&s->pb, 1, 1); s->last_bits++; - s->misc_bits++; + s->misc_bits++; s->skip_count++; - return; - } + return; + } if (s->use_skip_mb_code) - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ if(s->msmpeg4_version<=2){ put_bits(&s->pb, @@ -599,10 +599,10 @@ } s->p_tex_bits += get_bits_diff(s); } else { - /* compute cbp */ - cbp = 0; + /* compute cbp */ + cbp = 0; coded_cbp = 0; - for (i = 0; i < 6; i++) { + for (i = 0; i < 6; i++) { int val, pred; val = (s->block_last_index[i] >= 1); cbp |= val << (5 - i); @@ -613,7 +613,7 @@ val = val ^ pred; } coded_cbp |= val << (5 - i); - } + } #if 0 if (coded_cbp) printf("cbp=%x %x\n", cbp, coded_cbp); @@ -625,12 +625,12 @@ v2_intra_cbpc[cbp&3][1], v2_intra_cbpc[cbp&3][0]); } else { if (s->use_skip_mb_code) - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ put_bits(&s->pb, v2_mb_type[(cbp&3) + 4][1], v2_mb_type[(cbp&3) + 4][0]); } - put_bits(&s->pb, 1, 0); /* no AC prediction yet */ + put_bits(&s->pb, 1, 0); /* no AC prediction yet */ put_bits(&s->pb, cbpy_tab[cbp>>2][1], cbpy_tab[cbp>>2][0]); @@ -641,13 +641,13 @@ ff_msmp4_mb_i_table[coded_cbp][1], ff_msmp4_mb_i_table[coded_cbp][0]); } else { if (s->use_skip_mb_code) - put_bits(&s->pb, 1, 0); /* mb coded */ + put_bits(&s->pb, 1, 0); /* mb coded */ put_bits(&s->pb, table_mb_non_intra[cbp][1], table_mb_non_intra[cbp][0]); } set_stat(ST_INTRA_MB); - put_bits(&s->pb, 1, 0); /* no AC prediction yet */ + put_bits(&s->pb, 1, 0); /* no AC prediction yet */ if(s->inter_intra_pred){ s->h263_aic_dir=0; put_bits(&s->pb, table_inter_intra[s->h263_aic_dir][1], table_inter_intra[s->h263_aic_dir][0]); @@ -702,9 +702,9 @@ /* find prediction */ if (n < 4) { - scale = s->y_dc_scale; + scale = s->y_dc_scale; } else { - scale = s->c_dc_scale; + scale = s->c_dc_scale; } wrap = s->block_wrap[n]; @@ -727,22 +727,22 @@ to problems if Q could vary !) */ #if (defined(ARCH_X86) || defined(ARCH_X86_64)) && !defined PIC asm volatile( - "movl %3, %%eax \n\t" - "shrl $1, %%eax \n\t" - "addl %%eax, %2 \n\t" - "addl %%eax, %1 \n\t" - "addl %0, %%eax \n\t" - "mull %4 \n\t" - "movl %%edx, %0 \n\t" - "movl %1, %%eax \n\t" - "mull %4 \n\t" - "movl %%edx, %1 \n\t" - "movl %2, %%eax \n\t" - "mull %4 \n\t" - "movl %%edx, %2 \n\t" - : "+b" (a), "+c" (b), "+D" (c) - : "g" (scale), "S" (inverse[scale]) - : "%eax", "%edx" + "movl %3, %%eax \n\t" + "shrl $1, %%eax \n\t" + "addl %%eax, %2 \n\t" + "addl %%eax, %1 \n\t" + "addl %0, %%eax \n\t" + "mull %4 \n\t" + "movl %%edx, %0 \n\t" + "movl %1, %%eax \n\t" + "mull %4 \n\t" + "movl %%edx, %1 \n\t" + "movl %2, %%eax \n\t" + "mull %4 \n\t" + "movl %%edx, %2 \n\t" + : "+b" (a), "+c" (b), "+D" (c) + : "g" (scale), "S" (inverse[scale]) + : "%eax", "%edx" ); #else /* #elif defined (ARCH_ALPHA) */ @@ -750,13 +750,13 @@ common case. But they are costly everywhere... */ if (scale == 8) { - a = (a + (8 >> 1)) / 8; - b = (b + (8 >> 1)) / 8; - c = (c + (8 >> 1)) / 8; + a = (a + (8 >> 1)) / 8; + b = (b + (8 >> 1)) / 8; + c = (c + (8 >> 1)) / 8; } else { - a = FASTDIV((a + (scale >> 1)), scale); - b = FASTDIV((b + (scale >> 1)), scale); - c = FASTDIV((c + (scale >> 1)), scale); + a = FASTDIV((a + (scale >> 1)), scale); + b = FASTDIV((b + (scale >> 1)), scale); + c = FASTDIV((c + (scale >> 1)), scale); } #endif /* XXX: WARNING: they did not choose the same test as MPEG4. This @@ -957,17 +957,17 @@ /* AC coefs */ last_non_zero = i - 1; for (; i <= last_index; i++) { - j = scantable[i]; - level = block[j]; - if (level) { - run = i - last_non_zero - 1; - last = (i == last_index); - sign = 0; - slevel = level; - if (level < 0) { - sign = 1; - level = -level; - } + j = scantable[i]; + level = block[j]; + if (level) { + run = i - last_non_zero - 1; + last = (i == last_index); + sign = 0; + slevel = level; + if (level < 0) { + sign = 1; + level = -level; + } if(level<=MAX_LEVEL && run<=MAX_RUN){ s->ac_stats[s->mb_intra][n>3][level][run][last]++; @@ -1030,8 +1030,8 @@ } else { put_bits(&s->pb, 1, sign); } - last_non_zero = i; - } + last_non_zero = i; + } } } @@ -1064,7 +1064,7 @@ v = abs(level); while (v) { v >>= 1; - size++; + size++; } if (level < 0) @@ -1301,11 +1301,11 @@ } s->no_rounding = 1; if(s->avctx->debug&FF_DEBUG_PICT_INFO) - av_log(s->avctx, AV_LOG_DEBUG, "qscale:%d rlc:%d rl:%d dc:%d mbrl:%d slice:%d \n", - s->qscale, - s->rl_chroma_table_index, - s->rl_table_index, - s->dc_table_index, + av_log(s->avctx, AV_LOG_DEBUG, "qscale:%d rlc:%d rl:%d dc:%d mbrl:%d slice:%d \n", + s->qscale, + s->rl_chroma_table_index, + s->rl_table_index, + s->dc_table_index, s->per_mb_rl_table, s->slice_height); } else { @@ -1349,20 +1349,20 @@ } if(s->avctx->debug&FF_DEBUG_PICT_INFO) - av_log(s->avctx, AV_LOG_DEBUG, "skip:%d rl:%d rlc:%d dc:%d mv:%d mbrl:%d qp:%d \n", - s->use_skip_mb_code, - s->rl_table_index, - s->rl_chroma_table_index, - s->dc_table_index, - s->mv_table_index, + av_log(s->avctx, AV_LOG_DEBUG, "skip:%d rl:%d rlc:%d dc:%d mv:%d mbrl:%d qp:%d \n", + s->use_skip_mb_code, + s->rl_table_index, + s->rl_chroma_table_index, + s->dc_table_index, + s->mv_table_index, s->per_mb_rl_table, s->qscale); - if(s->flipflop_rounding){ - s->no_rounding ^= 1; - }else{ - s->no_rounding = 0; - } + if(s->flipflop_rounding){ + s->no_rounding ^= 1; + }else{ + s->no_rounding = 0; + } } //printf("%d %d %d %d %d\n", s->pict_type, s->bit_rate, s->inter_intra_pred, s->width, s->height); @@ -1557,10 +1557,10 @@ s->dsp.clear_blocks(s->block[0]); for (i = 0; i < 6; i++) { if (msmpeg4_decode_block(s, block[i], i, (cbp >> (5 - i)) & 1, NULL) < 0) - { + { av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); return -1; - } + } } return 0; } @@ -1593,8 +1593,8 @@ code = get_vlc2(&s->gb, mb_non_intra_vlc[DEFAULT_INTER_INDEX].table, MB_NON_INTRA_VLC_BITS, 3); if (code < 0) return -1; - //s->mb_intra = (code & 0x40) ? 0 : 1; - s->mb_intra = (~code & 0x40) >> 6; + //s->mb_intra = (code & 0x40) ? 0 : 1; + s->mb_intra = (~code & 0x40) >> 6; cbp = code & 0x3f; } else { @@ -1650,10 +1650,10 @@ s->dsp.clear_blocks(s->block[0]); for (i = 0; i < 6; i++) { if (msmpeg4_decode_block(s, block[i], i, (cbp >> (5 - i)) & 1, NULL) < 0) - { - av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); - return -1; - } + { + av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); + return -1; + } } return 0; @@ -1672,7 +1672,7 @@ qmul=1; qadd=0; - /* DC coef */ + /* DC coef */ set_stat(ST_DC); level = msmpeg4_decode_dc(s, n, &dc_pred_dir); @@ -1808,8 +1808,8 @@ } } #endif - //level = level * qmul + (level>0) * qadd - (level<=0) * qadd ; - if (level>0) level= level * qmul + qadd; + //level = level * qmul + (level>0) * qadd - (level<=0) * qadd ; + if (level>0) level= level * qmul + qadd; else level= level * qmul - qadd; #if 0 // waste of time too :( if(level>2048 || level<-2048){
--- a/oggtheora.c Wed Dec 21 17:50:40 2005 +0000 +++ b/oggtheora.c Thu Dec 22 01:10:11 2005 +0000 @@ -45,7 +45,7 @@ thc->op.bytes = buf_size; if(theora_decode_packetin(&thc->state, &thc->op)) - return -1; + return -1; theora_decode_YUVout(&thc->state, &yuv); @@ -78,7 +78,7 @@ uint8_t *cdp; if(ctx->extradata_size < 6) - return -1; + return -1; theora_info_init(&thc->info); @@ -87,25 +87,25 @@ size = ctx->extradata_size; for(i = 0; i < 3; i++){ - hs = *cdp++ << 8; - hs += *cdp++; - size -= 2; + hs = *cdp++ << 8; + hs += *cdp++; + size -= 2; - if(hs > size){ - av_log(ctx, AV_LOG_ERROR, "extradata too small: %i > %i\n", + if(hs > size){ + av_log(ctx, AV_LOG_ERROR, "extradata too small: %i > %i\n", hs, size); - return -1; - } + return -1; + } - op.packet = cdp; - op.bytes = hs; - op.b_o_s = !i; - if(theora_decode_header(&thc->info, &thc->comment, &op)) - return -1; - op.packetno++; + op.packet = cdp; + op.bytes = hs; + op.b_o_s = !i; + if(theora_decode_header(&thc->info, &thc->comment, &op)) + return -1; + op.packetno++; - cdp += hs; - size -= hs; + cdp += hs; + size -= hs; } theora_decode_init(&thc->state, &thc->info);
--- a/oggvorbis.c Wed Dec 21 17:50:40 2005 +0000 +++ b/oggvorbis.c Thu Dec 22 01:10:11 2005 +0000 @@ -40,13 +40,13 @@ return (vorbis_encode_setup_managed(vi, avccontext->channels, avccontext->sample_rate, -1, avccontext->bit_rate, -1) || - vorbis_encode_ctl(vi, OV_ECTL_RATEMANAGE_AVG, NULL) || - vorbis_encode_setup_init(vi)) ; + vorbis_encode_ctl(vi, OV_ECTL_RATEMANAGE_AVG, NULL) || + vorbis_encode_setup_init(vi)) ; #else /* constant bitrate */ return vorbis_encode_init(vi, avccontext->channels, - avccontext->sample_rate, -1, avccontext->bit_rate, -1) ; + avccontext->sample_rate, -1, avccontext->bit_rate, -1) ; #endif } @@ -58,8 +58,8 @@ vorbis_info_init(&context->vi) ; if(oggvorbis_init_encoder(&context->vi, avccontext) < 0) { - av_log(avccontext, AV_LOG_ERROR, "oggvorbis_encode_init: init_encoder failed") ; - return -1 ; + av_log(avccontext, AV_LOG_ERROR, "oggvorbis_encode_init: init_encoder failed") ; + return -1 ; } vorbis_analysis_init(&context->vd, &context->vi) ; vorbis_block_init(&context->vd, &context->vb) ; @@ -101,8 +101,8 @@ static int oggvorbis_encode_frame(AVCodecContext *avccontext, - unsigned char *packets, - int buf_size, void *data) + unsigned char *packets, + int buf_size, void *data) { OggVorbisContext *context = avccontext->priv_data ; float **buffer ; @@ -113,22 +113,22 @@ buffer = vorbis_analysis_buffer(&context->vd, samples) ; if(context->vi.channels == 1) { - for(l = 0 ; l < samples ; l++) - buffer[0][l]=audio[l]/32768.f; + for(l = 0 ; l < samples ; l++) + buffer[0][l]=audio[l]/32768.f; } else { - for(l = 0 ; l < samples ; l++){ - buffer[0][l]=audio[l*2]/32768.f; - buffer[1][l]=audio[l*2+1]/32768.f; - } + for(l = 0 ; l < samples ; l++){ + buffer[0][l]=audio[l*2]/32768.f; + buffer[1][l]=audio[l*2+1]/32768.f; + } } vorbis_analysis_wrote(&context->vd, samples) ; while(vorbis_analysis_blockout(&context->vd, &context->vb) == 1) { - vorbis_analysis(&context->vb, NULL); - vorbis_bitrate_addblock(&context->vb) ; + vorbis_analysis(&context->vb, NULL); + vorbis_bitrate_addblock(&context->vb) ; - while(vorbis_bitrate_flushpacket(&context->vd, &op)) { + while(vorbis_bitrate_flushpacket(&context->vd, &op)) { if(op.bytes==1) //id love to say this is a hack, bad sadly its not, appearently the end of stream decission is in libogg continue; memcpy(context->buffer + context->buffer_index, &op, sizeof(ogg_packet)); @@ -136,7 +136,7 @@ memcpy(context->buffer + context->buffer_index, op.packet, op.bytes); context->buffer_index += op.bytes; // av_log(avccontext, AV_LOG_DEBUG, "e%d / %d\n", context->buffer_index, op.bytes); - } + } } l=0; @@ -268,19 +268,19 @@ float *mono ; for(i = 0 ; i < channels ; i++){ - ptr = &data[i]; - mono = pcm[i] ; + ptr = &data[i]; + mono = pcm[i] ; - for(j = 0 ; j < samples ; j++) { + for(j = 0 ; j < samples ; j++) { - val = mono[j] * 32767.f; + val = mono[j] * 32767.f; - if(val > 32767) val = 32767 ; - if(val < -32768) val = -32768 ; + if(val > 32767) val = 32767 ; + if(val < -32768) val = -32768 ; - *ptr = val ; - ptr += channels; - } + *ptr = val ; + ptr += channels; + } } return 0 ; @@ -311,15 +311,15 @@ av_log(avccontext, AV_LOG_DEBUG, "\n");*/ if(vorbis_synthesis(&context->vb, op) == 0) - vorbis_synthesis_blockin(&context->vd, &context->vb) ; + vorbis_synthesis_blockin(&context->vd, &context->vb) ; total_samples = 0 ; total_bytes = 0 ; while((samples = vorbis_synthesis_pcmout(&context->vd, &pcm)) > 0) { - conv(samples, pcm, (char*)data + total_bytes, context->vi.channels) ; - total_bytes += samples * 2 * context->vi.channels ; - total_samples += samples ; + conv(samples, pcm, (char*)data + total_bytes, context->vi.channels) ; + total_bytes += samples * 2 * context->vi.channels ; + total_samples += samples ; vorbis_synthesis_read(&context->vd, samples) ; }
--- a/parser.c Wed Dec 21 17:50:40 2005 +0000 +++ b/parser.c Thu Dec 22 01:10:11 2005 +0000 @@ -191,11 +191,11 @@ //#define END_NOT_FOUND (-100) -#define PICTURE_START_CODE 0x00000100 -#define SEQ_START_CODE 0x000001b3 -#define EXT_START_CODE 0x000001b5 -#define SLICE_MIN_START_CODE 0x00000101 -#define SLICE_MAX_START_CODE 0x000001af +#define PICTURE_START_CODE 0x00000100 +#define SEQ_START_CODE 0x000001b3 +#define EXT_START_CODE 0x000001b5 +#define SLICE_MIN_START_CODE 0x00000101 +#define SLICE_MAX_START_CODE 0x000001af typedef struct ParseContext1{ ParseContext pc; @@ -571,7 +571,7 @@ /*************************/ typedef struct MpegAudioParseContext { - uint8_t inbuf[MPA_MAX_CODED_FRAME_SIZE]; /* input buffer */ + uint8_t inbuf[MPA_MAX_CODED_FRAME_SIZE]; /* input buffer */ uint8_t *inbuf_ptr; int frame_size; int free_format_frame_size; @@ -608,8 +608,8 @@ *poutbuf_size = 0; buf_ptr = buf; while (buf_size > 0) { - len = s->inbuf_ptr - s->inbuf; - if (s->frame_size == 0) { + len = s->inbuf_ptr - s->inbuf; + if (s->frame_size == 0) { /* special case for next header for first frame in free format case (XXX: find a simpler method) */ if (s->free_format_next_header != 0) { @@ -621,34 +621,34 @@ s->free_format_next_header = 0; goto got_header; } - /* no header seen : find one. We need at least MPA_HEADER_SIZE + /* no header seen : find one. We need at least MPA_HEADER_SIZE bytes to parse it */ - len = MPA_HEADER_SIZE - len; - if (len > buf_size) - len = buf_size; - if (len > 0) { - memcpy(s->inbuf_ptr, buf_ptr, len); - buf_ptr += len; - buf_size -= len; - s->inbuf_ptr += len; - } - if ((s->inbuf_ptr - s->inbuf) >= MPA_HEADER_SIZE) { + len = MPA_HEADER_SIZE - len; + if (len > buf_size) + len = buf_size; + if (len > 0) { + memcpy(s->inbuf_ptr, buf_ptr, len); + buf_ptr += len; + buf_size -= len; + s->inbuf_ptr += len; + } + if ((s->inbuf_ptr - s->inbuf) >= MPA_HEADER_SIZE) { got_header: sr= avctx->sample_rate; - header = (s->inbuf[0] << 24) | (s->inbuf[1] << 16) | - (s->inbuf[2] << 8) | s->inbuf[3]; + header = (s->inbuf[0] << 24) | (s->inbuf[1] << 16) | + (s->inbuf[2] << 8) | s->inbuf[3]; ret = mpa_decode_header(avctx, header); if (ret < 0) { s->header_count= -2; - /* no sync found : move by one byte (inefficient, but simple!) */ - memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); - s->inbuf_ptr--; + /* no sync found : move by one byte (inefficient, but simple!) */ + memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); + s->inbuf_ptr--; dprintf("skip %x\n", header); /* reset free format frame size to give a chance to get a new bitrate */ s->free_format_frame_size = 0; - } else { + } else { if((header&SAME_HEADER_MASK) != (s->header&SAME_HEADER_MASK) && s->header) s->header_count= -3; s->header= header; @@ -657,26 +657,26 @@ #if 0 /* free format: prepare to compute frame size */ - if (decode_header(s, header) == 1) { - s->frame_size = -1; + if (decode_header(s, header) == 1) { + s->frame_size = -1; } #endif - } + } if(s->header_count <= 0) avctx->sample_rate= sr; //FIXME ugly - } + } } else #if 0 if (s->frame_size == -1) { /* free format : find next sync to compute frame size */ - len = MPA_MAX_CODED_FRAME_SIZE - len; - if (len > buf_size) - len = buf_size; + len = MPA_MAX_CODED_FRAME_SIZE - len; + if (len > buf_size) + len = buf_size; if (len == 0) { - /* frame too long: resync */ + /* frame too long: resync */ s->frame_size = 0; - memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); - s->inbuf_ptr--; + memmove(s->inbuf, s->inbuf + 1, s->inbuf_ptr - s->inbuf - 1); + s->inbuf_ptr--; } else { uint8_t *p, *pend; uint32_t header1; @@ -720,19 +720,19 @@ s->inbuf_ptr += len; buf_size -= len; } - } else + } else #endif if (len < s->frame_size) { if (s->frame_size > MPA_MAX_CODED_FRAME_SIZE) s->frame_size = MPA_MAX_CODED_FRAME_SIZE; - len = s->frame_size - len; - if (len > buf_size) - len = buf_size; - memcpy(s->inbuf_ptr, buf_ptr, len); - buf_ptr += len; - s->inbuf_ptr += len; - buf_size -= len; - } + len = s->frame_size - len; + if (len > buf_size) + len = buf_size; + memcpy(s->inbuf_ptr, buf_ptr, len); + buf_ptr += len; + s->inbuf_ptr += len; + buf_size -= len; + } // next_data: if (s->frame_size > 0 && (s->inbuf_ptr - s->inbuf) >= s->frame_size) { @@ -740,10 +740,10 @@ *poutbuf = s->inbuf; *poutbuf_size = s->inbuf_ptr - s->inbuf; } - s->inbuf_ptr = s->inbuf; - s->frame_size = 0; - break; - } + s->inbuf_ptr = s->inbuf; + s->frame_size = 0; + break; + } } return buf_ptr - buf; } @@ -783,7 +783,7 @@ const uint8_t *buf_ptr; int len, sample_rate, bit_rate; static const int ac3_channels[8] = { - 2, 1, 2, 3, 3, 4, 4, 5 + 2, 1, 2, 3, 3, 4, 4, 5 }; *poutbuf = NULL; @@ -812,7 +812,7 @@ memmove(s->inbuf, s->inbuf + 1, AC3_HEADER_SIZE - 1); s->inbuf_ptr--; } else { - s->frame_size = len; + s->frame_size = len; /* update codec info */ avctx->sample_rate = sample_rate; /* set channels,except if the user explicitly requests 1 or 2 channels, XXX/FIXME this is a bit ugly */ @@ -821,7 +821,7 @@ if (s->flags & A52_LFE) avctx->channels++; } - avctx->bit_rate = bit_rate; + avctx->bit_rate = bit_rate; avctx->frame_size = 6 * 256; } }
--- a/pcm.c Wed Dec 21 17:50:40 2005 +0000 +++ b/pcm.c Thu Dec 22 01:10:11 2005 +0000 @@ -27,48 +27,48 @@ /* from g711.c by SUN microsystems (unrestricted use) */ -#define SIGN_BIT (0x80) /* Sign bit for a A-law byte. */ -#define QUANT_MASK (0xf) /* Quantization field mask. */ -#define NSEGS (8) /* Number of A-law segments. */ -#define SEG_SHIFT (4) /* Left shift for segment number. */ -#define SEG_MASK (0x70) /* Segment field mask. */ +#define SIGN_BIT (0x80) /* Sign bit for a A-law byte. */ +#define QUANT_MASK (0xf) /* Quantization field mask. */ +#define NSEGS (8) /* Number of A-law segments. */ +#define SEG_SHIFT (4) /* Left shift for segment number. */ +#define SEG_MASK (0x70) /* Segment field mask. */ -#define BIAS (0x84) /* Bias for linear code. */ +#define BIAS (0x84) /* Bias for linear code. */ /* * alaw2linear() - Convert an A-law value to 16-bit linear PCM * */ -static int alaw2linear(unsigned char a_val) +static int alaw2linear(unsigned char a_val) { - int t; - int seg; + int t; + int seg; - a_val ^= 0x55; + a_val ^= 0x55; - t = a_val & QUANT_MASK; - seg = ((unsigned)a_val & SEG_MASK) >> SEG_SHIFT; - if(seg) t= (t + t + 1 + 32) << (seg + 2); - else t= (t + t + 1 ) << 3; + t = a_val & QUANT_MASK; + seg = ((unsigned)a_val & SEG_MASK) >> SEG_SHIFT; + if(seg) t= (t + t + 1 + 32) << (seg + 2); + else t= (t + t + 1 ) << 3; - return ((a_val & SIGN_BIT) ? t : -t); + return ((a_val & SIGN_BIT) ? t : -t); } -static int ulaw2linear(unsigned char u_val) +static int ulaw2linear(unsigned char u_val) { - int t; + int t; - /* Complement to obtain normal u-law value. */ - u_val = ~u_val; + /* Complement to obtain normal u-law value. */ + u_val = ~u_val; - /* - * Extract and bias the quantization bits. Then - * shift up by the segment number and subtract out the bias. - */ - t = ((u_val & QUANT_MASK) << 3) + BIAS; - t <<= ((unsigned)u_val & SEG_MASK) >> SEG_SHIFT; + /* + * Extract and bias the quantization bits. Then + * shift up by the segment number and subtract out the bias. + */ + t = ((u_val & QUANT_MASK) << 3) + BIAS; + t <<= ((unsigned)u_val & SEG_MASK) >> SEG_SHIFT; - return ((u_val & SIGN_BIT) ? (BIAS - t) : (t - BIAS)); + return ((u_val & SIGN_BIT) ? (BIAS - t) : (t - BIAS)); } /* 16384 entries per table */ @@ -209,7 +209,7 @@ } static int pcm_encode_frame(AVCodecContext *avctx, - unsigned char *frame, int buf_size, void *data) + unsigned char *frame, int buf_size, void *data) { int n, sample_size, v; short *samples; @@ -397,8 +397,8 @@ } static int pcm_decode_frame(AVCodecContext *avctx, - void *data, int *data_size, - uint8_t *buf, int buf_size) + void *data, int *data_size, + uint8_t *buf, int buf_size) { PCMDecode *s = avctx->priv_data; int n; @@ -509,9 +509,9 @@ CODEC_TYPE_AUDIO, \ id, \ 0, \ - pcm_encode_init, \ - pcm_encode_frame, \ - pcm_encode_close, \ + pcm_encode_init, \ + pcm_encode_frame, \ + pcm_encode_close, \ NULL, \ }; \ AVCodec name ## _decoder = { \ @@ -519,7 +519,7 @@ CODEC_TYPE_AUDIO, \ id, \ sizeof(PCMDecode), \ - pcm_decode_init, \ + pcm_decode_init, \ NULL, \ NULL, \ pcm_decode_frame, \
--- a/ppc/dsputil_altivec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/dsputil_altivec.c Thu Dec 22 01:10:11 2005 +0000 @@ -67,7 +67,7 @@ /* Read unaligned pixels into our vectors. The vectors are as follows: pix1v: pix1[0]-pix1[15] - pix2v: pix2[0]-pix2[15] pix2iv: pix2[1]-pix2[16] + pix2v: pix2[0]-pix2[15] pix2iv: pix2[1]-pix2[16] */ tv = (vector unsigned char *) pix1; pix1v = vec_perm(tv[0], tv[1], vec_lvsl(0, pix1)); @@ -184,7 +184,7 @@ fact to avoid a potentially expensive unaligned read, as well as some splitting, and vector addition each time around the loop. Read unaligned pixels into our vectors. The vectors are as follows: - pix2v: pix2[0]-pix2[15] pix2iv: pix2[1]-pix2[16] + pix2v: pix2[0]-pix2[15] pix2iv: pix2[1]-pix2[16] Split the pixel vectors into shorts */ tv = (vector unsigned char *) &pix2[0]; @@ -204,7 +204,7 @@ /* Read unaligned pixels into our vectors. The vectors are as follows: pix1v: pix1[0]-pix1[15] - pix3v: pix3[0]-pix3[15] pix3iv: pix3[1]-pix3[16] + pix3v: pix3[0]-pix3[15] pix3iv: pix3[1]-pix3[16] */ tv = (vector unsigned char *) pix1; pix1v = vec_perm(tv[0], tv[1], vec_lvsl(0, pix1)); @@ -273,7 +273,7 @@ for(i=0;i<h;i++) { - /* Read potentially unaligned pixels into t1 and t2 */ + /* Read potentially unaligned pixels into t1 and t2 */ perm1 = vec_lvsl(0, pix1); pix1v = (vector unsigned char *) pix1; perm2 = vec_lvsl(0, pix2); @@ -281,12 +281,12 @@ t1 = vec_perm(pix1v[0], pix1v[1], perm1); t2 = vec_perm(pix2v[0], pix2v[1], perm2); - /* Calculate a sum of abs differences vector */ + /* Calculate a sum of abs differences vector */ t3 = vec_max(t1, t2); t4 = vec_min(t1, t2); t5 = vec_sub(t3, t4); - /* Add each 4 pixel group together and put 4 results into sad */ + /* Add each 4 pixel group together and put 4 results into sad */ sad = vec_sum4s(t5, sad); pix1 += line_size; @@ -316,9 +316,9 @@ permclear = (vector unsigned char)AVV(255,255,255,255,255,255,255,255,0,0,0,0,0,0,0,0); for(i=0;i<h;i++) { - /* Read potentially unaligned pixels into t1 and t2 - Since we're reading 16 pixels, and actually only want 8, - mask out the last 8 pixels. The 0s don't change the sum. */ + /* Read potentially unaligned pixels into t1 and t2 + Since we're reading 16 pixels, and actually only want 8, + mask out the last 8 pixels. The 0s don't change the sum. */ perm1 = vec_lvsl(0, pix1); pix1v = (vector unsigned char *) pix1; perm2 = vec_lvsl(0, pix2); @@ -326,12 +326,12 @@ t1 = vec_and(vec_perm(pix1v[0], pix1v[1], perm1), permclear); t2 = vec_and(vec_perm(pix2v[0], pix2v[1], perm2), permclear); - /* Calculate a sum of abs differences vector */ + /* Calculate a sum of abs differences vector */ t3 = vec_max(t1, t2); t4 = vec_min(t1, t2); t5 = vec_sub(t3, t4); - /* Add each 4 pixel group together and put 4 results into sad */ + /* Add each 4 pixel group together and put 4 results into sad */ sad = vec_sum4s(t5, sad); pix1 += line_size; @@ -398,9 +398,9 @@ for(i=0;i<h;i++) { - /* Read potentially unaligned pixels into t1 and t2 - Since we're reading 16 pixels, and actually only want 8, - mask out the last 8 pixels. The 0s don't change the sum. */ + /* Read potentially unaligned pixels into t1 and t2 + Since we're reading 16 pixels, and actually only want 8, + mask out the last 8 pixels. The 0s don't change the sum. */ perm1 = vec_lvsl(0, pix1); pix1v = (vector unsigned char *) pix1; perm2 = vec_lvsl(0, pix2); @@ -413,7 +413,7 @@ of the fact that abs(a-b)^2 = (a-b)^2. */ - /* Calculate abs differences vector */ + /* Calculate abs differences vector */ t3 = vec_max(t1, t2); t4 = vec_min(t1, t2); t5 = vec_sub(t3, t4); @@ -451,7 +451,7 @@ sum = (vector unsigned int)vec_splat_u32(0); for(i=0;i<h;i++) { - /* Read potentially unaligned pixels into t1 and t2 */ + /* Read potentially unaligned pixels into t1 and t2 */ perm1 = vec_lvsl(0, pix1); pix1v = (vector unsigned char *) pix1; perm2 = vec_lvsl(0, pix2); @@ -464,7 +464,7 @@ of the fact that abs(a-b)^2 = (a-b)^2. */ - /* Calculate abs differences vector */ + /* Calculate abs differences vector */ t3 = vec_max(t1, t2); t4 = vec_min(t1, t2); t5 = vec_sub(t3, t4); @@ -498,12 +498,12 @@ sad = (vector unsigned int)vec_splat_u32(0); for (i = 0; i < 16; i++) { - /* Read the potentially unaligned 16 pixels into t1 */ + /* Read the potentially unaligned 16 pixels into t1 */ perm = vec_lvsl(0, pix); pixv = (vector unsigned char *) pix; t1 = vec_perm(pixv[0], pixv[1], perm); - /* Add each 4 pixel group together and put 4 results into sad */ + /* Add each 4 pixel group together and put 4 results into sad */ sad = vec_sum4s(t1, sad); pix += line_size; @@ -1335,32 +1335,32 @@ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07); -#define ONEITERBUTTERFLY(i, res) \ - { \ - register vector unsigned char src1, src2, srcO; \ - register vector unsigned char dst1, dst2, dstO; \ - src1 = vec_ld(stride * i, src); \ - if ((((stride * i) + (unsigned long)src) & 0x0000000F) > 8) \ - src2 = vec_ld((stride * i) + 16, src); \ - srcO = vec_perm(src1, src2, vec_lvsl(stride * i, src)); \ - dst1 = vec_ld(stride * i, dst); \ - if ((((stride * i) + (unsigned long)dst) & 0x0000000F) > 8) \ - dst2 = vec_ld((stride * i) + 16, dst); \ - dstO = vec_perm(dst1, dst2, vec_lvsl(stride * i, dst)); \ - /* promote the unsigned chars to signed shorts */ \ - /* we're in the 8x8 function, we only care for the first 8 */ \ - register vector signed short srcV = \ - (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)srcO); \ - register vector signed short dstV = \ - (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)dstO); \ - /* substractions inside the first butterfly */ \ - register vector signed short but0 = vec_sub(srcV, dstV); \ - register vector signed short op1 = vec_perm(but0, but0, perm1); \ - register vector signed short but1 = vec_mladd(but0, vprod1, op1); \ - register vector signed short op2 = vec_perm(but1, but1, perm2); \ - register vector signed short but2 = vec_mladd(but1, vprod2, op2); \ - register vector signed short op3 = vec_perm(but2, but2, perm3); \ - res = vec_mladd(but2, vprod3, op3); \ +#define ONEITERBUTTERFLY(i, res) \ + { \ + register vector unsigned char src1, src2, srcO; \ + register vector unsigned char dst1, dst2, dstO; \ + src1 = vec_ld(stride * i, src); \ + if ((((stride * i) + (unsigned long)src) & 0x0000000F) > 8) \ + src2 = vec_ld((stride * i) + 16, src); \ + srcO = vec_perm(src1, src2, vec_lvsl(stride * i, src)); \ + dst1 = vec_ld(stride * i, dst); \ + if ((((stride * i) + (unsigned long)dst) & 0x0000000F) > 8) \ + dst2 = vec_ld((stride * i) + 16, dst); \ + dstO = vec_perm(dst1, dst2, vec_lvsl(stride * i, dst)); \ + /* promote the unsigned chars to signed shorts */ \ + /* we're in the 8x8 function, we only care for the first 8 */ \ + register vector signed short srcV = \ + (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)srcO); \ + register vector signed short dstV = \ + (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)dstO); \ + /* substractions inside the first butterfly */ \ + register vector signed short but0 = vec_sub(srcV, dstV); \ + register vector signed short op1 = vec_perm(but0, but0, perm1); \ + register vector signed short but1 = vec_mladd(but0, vprod1, op1); \ + register vector signed short op2 = vec_perm(but1, but1, perm2); \ + register vector signed short but2 = vec_mladd(but1, vprod2, op2); \ + register vector signed short op3 = vec_perm(but2, but2, perm3); \ + res = vec_mladd(but2, vprod3, op3); \ } ONEITERBUTTERFLY(0, temp0); ONEITERBUTTERFLY(1, temp1); @@ -1480,26 +1480,26 @@ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07); -#define ONEITERBUTTERFLY(i, res1, res2) \ - { \ +#define ONEITERBUTTERFLY(i, res1, res2) \ + { \ register vector unsigned char src1 asm ("v22"), src2 asm ("v23"); \ register vector unsigned char dst1 asm ("v24"), dst2 asm ("v25"); \ - src1 = vec_ld(stride * i, src); \ - src2 = vec_ld((stride * i) + 16, src); \ + src1 = vec_ld(stride * i, src); \ + src2 = vec_ld((stride * i) + 16, src); \ register vector unsigned char srcO asm ("v22") = vec_perm(src1, src2, vec_lvsl(stride * i, src)); \ - dst1 = vec_ld(stride * i, dst); \ - dst2 = vec_ld((stride * i) + 16, dst); \ + dst1 = vec_ld(stride * i, dst); \ + dst2 = vec_ld((stride * i) + 16, dst); \ register vector unsigned char dstO asm ("v23") = vec_perm(dst1, dst2, vec_lvsl(stride * i, dst)); \ - /* promote the unsigned chars to signed shorts */ \ + /* promote the unsigned chars to signed shorts */ \ register vector signed short srcV asm ("v24") = \ - (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)srcO); \ + (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)srcO); \ register vector signed short dstV asm ("v25") = \ - (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)dstO); \ + (vector signed short)vec_mergeh((vector signed char)vzero, (vector signed char)dstO); \ register vector signed short srcW asm ("v26") = \ - (vector signed short)vec_mergel((vector signed char)vzero, (vector signed char)srcO); \ + (vector signed short)vec_mergel((vector signed char)vzero, (vector signed char)srcO); \ register vector signed short dstW asm ("v27") = \ - (vector signed short)vec_mergel((vector signed char)vzero, (vector signed char)dstO); \ - /* substractions inside the first butterfly */ \ + (vector signed short)vec_mergel((vector signed char)vzero, (vector signed char)dstO); \ + /* substractions inside the first butterfly */ \ register vector signed short but0 asm ("v28") = vec_sub(srcV, dstV); \ register vector signed short but0S asm ("v29") = vec_sub(srcW, dstW); \ register vector signed short op1 asm ("v30") = vec_perm(but0, but0, perm1); \ @@ -1511,9 +1511,9 @@ register vector signed short op2S asm ("v27") = vec_perm(but1S, but1S, perm2); \ register vector signed short but2S asm ("v28") = vec_mladd(but1S, vprod2, op2S); \ register vector signed short op3 asm ("v29") = vec_perm(but2, but2, perm3); \ - res1 = vec_mladd(but2, vprod3, op3); \ + res1 = vec_mladd(but2, vprod3, op3); \ register vector signed short op3S asm ("v30") = vec_perm(but2S, but2S, perm3); \ - res2 = vec_mladd(but2S, vprod3, op3S); \ + res2 = vec_mladd(but2S, vprod3, op3S); \ } ONEITERBUTTERFLY(0, temp0, temp0S); ONEITERBUTTERFLY(1, temp1, temp1S); @@ -1623,12 +1623,12 @@ int has_altivec(void) { #ifdef __AMIGAOS4__ - ULONG result = 0; - extern struct ExecIFace *IExec; + ULONG result = 0; + extern struct ExecIFace *IExec; - IExec->GetCPUInfoTags(GCIT_VectorUnit, &result, TAG_DONE); - if (result == VECTORTYPE_ALTIVEC) return 1; - return 0; + IExec->GetCPUInfoTags(GCIT_VectorUnit, &result, TAG_DONE); + if (result == VECTORTYPE_ALTIVEC) return 1; + return 0; #else /* __AMIGAOS4__ */ #ifdef CONFIG_DARWIN
--- a/ppc/dsputil_h264_altivec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/dsputil_h264_altivec.c Thu Dec 22 01:10:11 2005 +0000 @@ -191,33 +191,33 @@ /* from dsputil.c */ static inline void put_pixels8_l2(uint8_t * dst, const uint8_t * src1, const uint8_t * src2, int dst_stride, int src_stride1, int src_stride2, int h) { - int i; - for (i = 0; i < h; i++) { - uint32_t a, b; - a = (((const struct unaligned_32 *) (&src1[i * src_stride1]))->l); - b = (((const struct unaligned_32 *) (&src2[i * src_stride2]))->l); - *((uint32_t *) & dst[i * dst_stride]) = rnd_avg32(a, b); - a = (((const struct unaligned_32 *) (&src1[i * src_stride1 + 4]))->l); - b = (((const struct unaligned_32 *) (&src2[i * src_stride2 + 4]))->l); - *((uint32_t *) & dst[i * dst_stride + 4]) = rnd_avg32(a, b); - } + int i; + for (i = 0; i < h; i++) { + uint32_t a, b; + a = (((const struct unaligned_32 *) (&src1[i * src_stride1]))->l); + b = (((const struct unaligned_32 *) (&src2[i * src_stride2]))->l); + *((uint32_t *) & dst[i * dst_stride]) = rnd_avg32(a, b); + a = (((const struct unaligned_32 *) (&src1[i * src_stride1 + 4]))->l); + b = (((const struct unaligned_32 *) (&src2[i * src_stride2 + 4]))->l); + *((uint32_t *) & dst[i * dst_stride + 4]) = rnd_avg32(a, b); + } } static inline void avg_pixels8_l2(uint8_t * dst, const uint8_t * src1, const uint8_t * src2, int dst_stride, int src_stride1, int src_stride2, int h) { - int i; - for (i = 0; i < h; i++) { - uint32_t a, b; - a = (((const struct unaligned_32 *) (&src1[i * src_stride1]))->l); - b = (((const struct unaligned_32 *) (&src2[i * src_stride2]))->l); - *((uint32_t *) & dst[i * dst_stride]) = rnd_avg32(*((uint32_t *) & dst[i * dst_stride]), rnd_avg32(a, b)); - a = (((const struct unaligned_32 *) (&src1[i * src_stride1 + 4]))->l); - b = (((const struct unaligned_32 *) (&src2[i * src_stride2 + 4]))->l); - *((uint32_t *) & dst[i * dst_stride + 4]) = rnd_avg32(*((uint32_t *) & dst[i * dst_stride + 4]), rnd_avg32(a, b)); - } + int i; + for (i = 0; i < h; i++) { + uint32_t a, b; + a = (((const struct unaligned_32 *) (&src1[i * src_stride1]))->l); + b = (((const struct unaligned_32 *) (&src2[i * src_stride2]))->l); + *((uint32_t *) & dst[i * dst_stride]) = rnd_avg32(*((uint32_t *) & dst[i * dst_stride]), rnd_avg32(a, b)); + a = (((const struct unaligned_32 *) (&src1[i * src_stride1 + 4]))->l); + b = (((const struct unaligned_32 *) (&src2[i * src_stride2 + 4]))->l); + *((uint32_t *) & dst[i * dst_stride + 4]) = rnd_avg32(*((uint32_t *) & dst[i * dst_stride + 4]), rnd_avg32(a, b)); + } } static inline void put_pixels16_l2(uint8_t * dst, const uint8_t * src1, const uint8_t * src2, int dst_stride, int src_stride1, int src_stride2, int h) { - put_pixels8_l2(dst, src1, src2, dst_stride, src_stride1, src_stride2, h); - put_pixels8_l2(dst + 8, src1 + 8, src2 + 8, dst_stride, src_stride1, src_stride2, h); + put_pixels8_l2(dst, src1, src2, dst_stride, src_stride1, src_stride2, h); + put_pixels8_l2(dst + 8, src1 + 8, src2 + 8, dst_stride, src_stride1, src_stride2, h); } static inline void avg_pixels16_l2(uint8_t * dst, const uint8_t * src1, const uint8_t * src2, int dst_stride, int src_stride1, int src_stride2, int h) { - avg_pixels8_l2(dst, src1, src2, dst_stride, src_stride1, src_stride2, h); - avg_pixels8_l2(dst + 8, src1 + 8, src2 + 8, dst_stride, src_stride1, src_stride2, h); + avg_pixels8_l2(dst, src1, src2, dst_stride, src_stride1, src_stride2, h); + avg_pixels8_l2(dst + 8, src1 + 8, src2 + 8, dst_stride, src_stride1, src_stride2, h); } /* UNIMPLEMENTED YET !! */
--- a/ppc/dsputil_ppc.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/dsputil_ppc.c Thu Dec 22 01:10:11 2005 +0000 @@ -87,16 +87,16 @@ { for (j = 0; j < POWERPC_NUM_PMC_ENABLED ; j++) { - if (perfdata[j][i][powerpc_data_num] != (unsigned long long)0) - av_log(NULL, AV_LOG_INFO, - " Function \"%s\" (pmc%d):\n\tmin: %llu\n\tmax: %llu\n\tavg: %1.2lf (%llu)\n", - perfname[i], - j+1, - perfdata[j][i][powerpc_data_min], - perfdata[j][i][powerpc_data_max], - (double)perfdata[j][i][powerpc_data_sum] / - (double)perfdata[j][i][powerpc_data_num], - perfdata[j][i][powerpc_data_num]); + if (perfdata[j][i][powerpc_data_num] != (unsigned long long)0) + av_log(NULL, AV_LOG_INFO, + " Function \"%s\" (pmc%d):\n\tmin: %llu\n\tmax: %llu\n\tavg: %1.2lf (%llu)\n", + perfname[i], + j+1, + perfdata[j][i][powerpc_data_min], + perfdata[j][i][powerpc_data_max], + (double)perfdata[j][i][powerpc_data_sum] / + (double)perfdata[j][i][powerpc_data_num], + perfdata[j][i][powerpc_data_num]); } } } @@ -179,7 +179,7 @@ } else for ( ; i < sizeof(DCTELEM)*6*64 ; i += 128) { - asm volatile("dcbzl %0,%1" : : "b" (blocks), "r" (i) : "memory"); + asm volatile("dcbzl %0,%1" : : "b" (blocks), "r" (i) : "memory"); } #else memset(blocks, 0, sizeof(DCTELEM)*6*64); @@ -284,25 +284,25 @@ c->put_no_rnd_pixels_tab[0][0] = put_pixels16_altivec; c->avg_pixels_tab[0][0] = avg_pixels16_altivec; c->avg_pixels_tab[1][0] = avg_pixels8_altivec; - c->avg_pixels_tab[1][3] = avg_pixels8_xy2_altivec; + c->avg_pixels_tab[1][3] = avg_pixels8_xy2_altivec; c->put_pixels_tab[1][3] = put_pixels8_xy2_altivec; c->put_no_rnd_pixels_tab[1][3] = put_no_rnd_pixels8_xy2_altivec; c->put_pixels_tab[0][3] = put_pixels16_xy2_altivec; c->put_no_rnd_pixels_tab[0][3] = put_no_rnd_pixels16_xy2_altivec; - c->gmc1 = gmc1_altivec; + c->gmc1 = gmc1_altivec; #ifdef CONFIG_DARWIN // ATM gcc-3.3 and gcc-3.4 fail to compile these in linux... - c->hadamard8_diff[0] = hadamard8_diff16_altivec; - c->hadamard8_diff[1] = hadamard8_diff8x8_altivec; + c->hadamard8_diff[0] = hadamard8_diff16_altivec; + c->hadamard8_diff[1] = hadamard8_diff8x8_altivec; #endif #ifdef CONFIG_ENCODERS - if (avctx->dct_algo == FF_DCT_AUTO || - avctx->dct_algo == FF_DCT_ALTIVEC) - { - c->fdct = fdct_altivec; - } + if (avctx->dct_algo == FF_DCT_AUTO || + avctx->dct_algo == FF_DCT_ALTIVEC) + { + c->fdct = fdct_altivec; + } #endif //CONFIG_ENCODERS if (avctx->lowres==0) @@ -325,14 +325,14 @@ int i, j; for (i = 0 ; i < powerpc_perf_total ; i++) { - for (j = 0; j < POWERPC_NUM_PMC_ENABLED ; j++) - { - perfdata[j][i][powerpc_data_min] = 0xFFFFFFFFFFFFFFFFULL; - perfdata[j][i][powerpc_data_max] = 0x0000000000000000ULL; - perfdata[j][i][powerpc_data_sum] = 0x0000000000000000ULL; - perfdata[j][i][powerpc_data_num] = 0x0000000000000000ULL; - } - } + for (j = 0; j < POWERPC_NUM_PMC_ENABLED ; j++) + { + perfdata[j][i][powerpc_data_min] = 0xFFFFFFFFFFFFFFFFULL; + perfdata[j][i][powerpc_data_max] = 0x0000000000000000ULL; + perfdata[j][i][powerpc_data_sum] = 0x0000000000000000ULL; + perfdata[j][i][powerpc_data_num] = 0x0000000000000000ULL; + } + } } #endif /* POWERPC_PERFORMANCE_REPORT */ } else
--- a/ppc/dsputil_ppc.h Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/dsputil_ppc.h Thu Dec 22 01:10:11 2005 +0000 @@ -114,10 +114,10 @@ #define POWERPC_GET_PMC6(a) do {} while (0) #endif #endif /* POWERPC_MODE_64BITS */ -#define POWERPC_PERF_DECLARE(a, cond) \ - POWERP_PMC_DATATYPE \ - pmc_start[POWERPC_NUM_PMC_ENABLED], \ - pmc_stop[POWERPC_NUM_PMC_ENABLED], \ +#define POWERPC_PERF_DECLARE(a, cond) \ + POWERP_PMC_DATATYPE \ + pmc_start[POWERPC_NUM_PMC_ENABLED], \ + pmc_stop[POWERPC_NUM_PMC_ENABLED], \ pmc_loop_index; #define POWERPC_PERF_START_COUNT(a, cond) do { \ POWERPC_GET_PMC6(pmc_start[5]); \ @@ -141,8 +141,8 @@ pmc_loop_index++) \ { \ if (pmc_stop[pmc_loop_index] >= pmc_start[pmc_loop_index]) \ - { \ - POWERP_PMC_DATATYPE diff = \ + { \ + POWERP_PMC_DATATYPE diff = \ pmc_stop[pmc_loop_index] - pmc_start[pmc_loop_index]; \ if (diff < perfdata[pmc_loop_index][a][powerpc_data_min]) \ perfdata[pmc_loop_index][a][powerpc_data_min] = diff; \
--- a/ppc/fft_altivec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/fft_altivec.c Thu Dec 22 01:10:11 2005 +0000 @@ -65,8 +65,8 @@ POWERPC_PERF_DECLARE(altivec_fft_num, s->nbits >= 6); #ifdef ALTIVEC_USE_REFERENCE_C_CODE int ln = s->nbits; - int j, np, np2; - int nblocks, nloops; + int j, np, np2; + int nblocks, nloops; register FFTComplex *p, *q; FFTComplex *exptab = s->exptab; int l; @@ -147,8 +147,8 @@ #endif int ln = s->nbits; - int j, np, np2; - int nblocks, nloops; + int j, np, np2; + int nblocks, nloops; register FFTComplex *p, *q; FFTComplex *cptr, *cptr1; int k;
--- a/ppc/gcc_fixes.h Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/gcc_fixes.h Thu Dec 22 01:10:11 2005 +0000 @@ -30,31 +30,31 @@ */ static inline vector signed char ff_vmrglb (vector signed char const A, - vector signed char const B) + vector signed char const B) { static const vector unsigned char lowbyte = { - 0x08, 0x18, 0x09, 0x19, 0x0a, 0x1a, 0x0b, 0x1b, - 0x0c, 0x1c, 0x0d, 0x1d, 0x0e, 0x1e, 0x0f, 0x1f + 0x08, 0x18, 0x09, 0x19, 0x0a, 0x1a, 0x0b, 0x1b, + 0x0c, 0x1c, 0x0d, 0x1d, 0x0e, 0x1e, 0x0f, 0x1f }; return vec_perm (A, B, lowbyte); } static inline vector signed short ff_vmrglh (vector signed short const A, - vector signed short const B) + vector signed short const B) { static const vector unsigned char lowhalf = { - 0x08, 0x09, 0x18, 0x19, 0x0a, 0x0b, 0x1a, 0x1b, - 0x0c, 0x0d, 0x1c, 0x1d, 0x0e, 0x0f, 0x1e, 0x1f + 0x08, 0x09, 0x18, 0x19, 0x0a, 0x0b, 0x1a, 0x1b, + 0x0c, 0x0d, 0x1c, 0x1d, 0x0e, 0x0f, 0x1e, 0x1f }; return vec_perm (A, B, lowhalf); } static inline vector signed int ff_vmrglw (vector signed int const A, - vector signed int const B) + vector signed int const B) { static const vector unsigned char lowword = { - 0x08, 0x09, 0x0a, 0x0b, 0x18, 0x19, 0x1a, 0x1b, - 0x0c, 0x0d, 0x0e, 0x0f, 0x1c, 0x1d, 0x1e, 0x1f + 0x08, 0x09, 0x0a, 0x0b, 0x18, 0x19, 0x1a, 0x1b, + 0x0c, 0x0d, 0x0e, 0x0f, 0x1c, 0x1d, 0x1e, 0x1f }; return vec_perm (A, B, lowword); }
--- a/ppc/idct_altivec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/idct_altivec.c Thu Dec 22 01:10:11 2005 +0000 @@ -51,108 +51,108 @@ #define vector_s32_t vector signed int #define vector_u32_t vector unsigned int -#define IDCT_HALF \ - /* 1st stage */ \ - t1 = vec_mradds (a1, vx7, vx1 ); \ - t8 = vec_mradds (a1, vx1, vec_subs (zero, vx7)); \ - t7 = vec_mradds (a2, vx5, vx3); \ - t3 = vec_mradds (ma2, vx3, vx5); \ - \ - /* 2nd stage */ \ - t5 = vec_adds (vx0, vx4); \ - t0 = vec_subs (vx0, vx4); \ - t2 = vec_mradds (a0, vx6, vx2); \ - t4 = vec_mradds (a0, vx2, vec_subs (zero, vx6)); \ - t6 = vec_adds (t8, t3); \ - t3 = vec_subs (t8, t3); \ - t8 = vec_subs (t1, t7); \ - t1 = vec_adds (t1, t7); \ - \ - /* 3rd stage */ \ - t7 = vec_adds (t5, t2); \ - t2 = vec_subs (t5, t2); \ - t5 = vec_adds (t0, t4); \ - t0 = vec_subs (t0, t4); \ - t4 = vec_subs (t8, t3); \ - t3 = vec_adds (t8, t3); \ - \ - /* 4th stage */ \ - vy0 = vec_adds (t7, t1); \ - vy7 = vec_subs (t7, t1); \ - vy1 = vec_mradds (c4, t3, t5); \ - vy6 = vec_mradds (mc4, t3, t5); \ - vy2 = vec_mradds (c4, t4, t0); \ - vy5 = vec_mradds (mc4, t4, t0); \ - vy3 = vec_adds (t2, t6); \ +#define IDCT_HALF \ + /* 1st stage */ \ + t1 = vec_mradds (a1, vx7, vx1 ); \ + t8 = vec_mradds (a1, vx1, vec_subs (zero, vx7)); \ + t7 = vec_mradds (a2, vx5, vx3); \ + t3 = vec_mradds (ma2, vx3, vx5); \ + \ + /* 2nd stage */ \ + t5 = vec_adds (vx0, vx4); \ + t0 = vec_subs (vx0, vx4); \ + t2 = vec_mradds (a0, vx6, vx2); \ + t4 = vec_mradds (a0, vx2, vec_subs (zero, vx6)); \ + t6 = vec_adds (t8, t3); \ + t3 = vec_subs (t8, t3); \ + t8 = vec_subs (t1, t7); \ + t1 = vec_adds (t1, t7); \ + \ + /* 3rd stage */ \ + t7 = vec_adds (t5, t2); \ + t2 = vec_subs (t5, t2); \ + t5 = vec_adds (t0, t4); \ + t0 = vec_subs (t0, t4); \ + t4 = vec_subs (t8, t3); \ + t3 = vec_adds (t8, t3); \ + \ + /* 4th stage */ \ + vy0 = vec_adds (t7, t1); \ + vy7 = vec_subs (t7, t1); \ + vy1 = vec_mradds (c4, t3, t5); \ + vy6 = vec_mradds (mc4, t3, t5); \ + vy2 = vec_mradds (c4, t4, t0); \ + vy5 = vec_mradds (mc4, t4, t0); \ + vy3 = vec_adds (t2, t6); \ vy4 = vec_subs (t2, t6); -#define IDCT \ - vector_s16_t vx0, vx1, vx2, vx3, vx4, vx5, vx6, vx7; \ - vector_s16_t vy0, vy1, vy2, vy3, vy4, vy5, vy6, vy7; \ - vector_s16_t a0, a1, a2, ma2, c4, mc4, zero, bias; \ - vector_s16_t t0, t1, t2, t3, t4, t5, t6, t7, t8; \ - vector_u16_t shift; \ - \ - c4 = vec_splat (constants[0], 0); \ - a0 = vec_splat (constants[0], 1); \ - a1 = vec_splat (constants[0], 2); \ - a2 = vec_splat (constants[0], 3); \ - mc4 = vec_splat (constants[0], 4); \ - ma2 = vec_splat (constants[0], 5); \ - bias = (vector_s16_t)vec_splat ((vector_s32_t)constants[0], 3); \ - \ - zero = vec_splat_s16 (0); \ - shift = vec_splat_u16 (4); \ - \ - vx0 = vec_mradds (vec_sl (block[0], shift), constants[1], zero); \ - vx1 = vec_mradds (vec_sl (block[1], shift), constants[2], zero); \ - vx2 = vec_mradds (vec_sl (block[2], shift), constants[3], zero); \ - vx3 = vec_mradds (vec_sl (block[3], shift), constants[4], zero); \ - vx4 = vec_mradds (vec_sl (block[4], shift), constants[1], zero); \ - vx5 = vec_mradds (vec_sl (block[5], shift), constants[4], zero); \ - vx6 = vec_mradds (vec_sl (block[6], shift), constants[3], zero); \ - vx7 = vec_mradds (vec_sl (block[7], shift), constants[2], zero); \ - \ - IDCT_HALF \ - \ - vx0 = vec_mergeh (vy0, vy4); \ - vx1 = vec_mergel (vy0, vy4); \ - vx2 = vec_mergeh (vy1, vy5); \ - vx3 = vec_mergel (vy1, vy5); \ - vx4 = vec_mergeh (vy2, vy6); \ - vx5 = vec_mergel (vy2, vy6); \ - vx6 = vec_mergeh (vy3, vy7); \ - vx7 = vec_mergel (vy3, vy7); \ - \ - vy0 = vec_mergeh (vx0, vx4); \ - vy1 = vec_mergel (vx0, vx4); \ - vy2 = vec_mergeh (vx1, vx5); \ - vy3 = vec_mergel (vx1, vx5); \ - vy4 = vec_mergeh (vx2, vx6); \ - vy5 = vec_mergel (vx2, vx6); \ - vy6 = vec_mergeh (vx3, vx7); \ - vy7 = vec_mergel (vx3, vx7); \ - \ - vx0 = vec_adds (vec_mergeh (vy0, vy4), bias); \ - vx1 = vec_mergel (vy0, vy4); \ - vx2 = vec_mergeh (vy1, vy5); \ - vx3 = vec_mergel (vy1, vy5); \ - vx4 = vec_mergeh (vy2, vy6); \ - vx5 = vec_mergel (vy2, vy6); \ - vx6 = vec_mergeh (vy3, vy7); \ - vx7 = vec_mergel (vy3, vy7); \ - \ - IDCT_HALF \ - \ - shift = vec_splat_u16 (6); \ - vx0 = vec_sra (vy0, shift); \ - vx1 = vec_sra (vy1, shift); \ - vx2 = vec_sra (vy2, shift); \ - vx3 = vec_sra (vy3, shift); \ - vx4 = vec_sra (vy4, shift); \ - vx5 = vec_sra (vy5, shift); \ - vx6 = vec_sra (vy6, shift); \ +#define IDCT \ + vector_s16_t vx0, vx1, vx2, vx3, vx4, vx5, vx6, vx7; \ + vector_s16_t vy0, vy1, vy2, vy3, vy4, vy5, vy6, vy7; \ + vector_s16_t a0, a1, a2, ma2, c4, mc4, zero, bias; \ + vector_s16_t t0, t1, t2, t3, t4, t5, t6, t7, t8; \ + vector_u16_t shift; \ + \ + c4 = vec_splat (constants[0], 0); \ + a0 = vec_splat (constants[0], 1); \ + a1 = vec_splat (constants[0], 2); \ + a2 = vec_splat (constants[0], 3); \ + mc4 = vec_splat (constants[0], 4); \ + ma2 = vec_splat (constants[0], 5); \ + bias = (vector_s16_t)vec_splat ((vector_s32_t)constants[0], 3); \ + \ + zero = vec_splat_s16 (0); \ + shift = vec_splat_u16 (4); \ + \ + vx0 = vec_mradds (vec_sl (block[0], shift), constants[1], zero); \ + vx1 = vec_mradds (vec_sl (block[1], shift), constants[2], zero); \ + vx2 = vec_mradds (vec_sl (block[2], shift), constants[3], zero); \ + vx3 = vec_mradds (vec_sl (block[3], shift), constants[4], zero); \ + vx4 = vec_mradds (vec_sl (block[4], shift), constants[1], zero); \ + vx5 = vec_mradds (vec_sl (block[5], shift), constants[4], zero); \ + vx6 = vec_mradds (vec_sl (block[6], shift), constants[3], zero); \ + vx7 = vec_mradds (vec_sl (block[7], shift), constants[2], zero); \ + \ + IDCT_HALF \ + \ + vx0 = vec_mergeh (vy0, vy4); \ + vx1 = vec_mergel (vy0, vy4); \ + vx2 = vec_mergeh (vy1, vy5); \ + vx3 = vec_mergel (vy1, vy5); \ + vx4 = vec_mergeh (vy2, vy6); \ + vx5 = vec_mergel (vy2, vy6); \ + vx6 = vec_mergeh (vy3, vy7); \ + vx7 = vec_mergel (vy3, vy7); \ + \ + vy0 = vec_mergeh (vx0, vx4); \ + vy1 = vec_mergel (vx0, vx4); \ + vy2 = vec_mergeh (vx1, vx5); \ + vy3 = vec_mergel (vx1, vx5); \ + vy4 = vec_mergeh (vx2, vx6); \ + vy5 = vec_mergel (vx2, vx6); \ + vy6 = vec_mergeh (vx3, vx7); \ + vy7 = vec_mergel (vx3, vx7); \ + \ + vx0 = vec_adds (vec_mergeh (vy0, vy4), bias); \ + vx1 = vec_mergel (vy0, vy4); \ + vx2 = vec_mergeh (vy1, vy5); \ + vx3 = vec_mergel (vy1, vy5); \ + vx4 = vec_mergeh (vy2, vy6); \ + vx5 = vec_mergel (vy2, vy6); \ + vx6 = vec_mergeh (vy3, vy7); \ + vx7 = vec_mergel (vy3, vy7); \ + \ + IDCT_HALF \ + \ + shift = vec_splat_u16 (6); \ + vx0 = vec_sra (vy0, shift); \ + vx1 = vec_sra (vy1, shift); \ + vx2 = vec_sra (vy2, shift); \ + vx3 = vec_sra (vy3, shift); \ + vx4 = vec_sra (vy4, shift); \ + vx5 = vec_sra (vy5, shift); \ + vx6 = vec_sra (vy6, shift); \ vx7 = vec_sra (vy7, shift); @@ -180,18 +180,18 @@ #endif IDCT -#define COPY(dest,src) \ - tmp = vec_packsu (src, src); \ - vec_ste ((vector_u32_t)tmp, 0, (unsigned int *)dest); \ +#define COPY(dest,src) \ + tmp = vec_packsu (src, src); \ + vec_ste ((vector_u32_t)tmp, 0, (unsigned int *)dest); \ vec_ste ((vector_u32_t)tmp, 4, (unsigned int *)dest); - COPY (dest, vx0) dest += stride; - COPY (dest, vx1) dest += stride; - COPY (dest, vx2) dest += stride; - COPY (dest, vx3) dest += stride; - COPY (dest, vx4) dest += stride; - COPY (dest, vx5) dest += stride; - COPY (dest, vx6) dest += stride; + COPY (dest, vx0) dest += stride; + COPY (dest, vx1) dest += stride; + COPY (dest, vx2) dest += stride; + COPY (dest, vx3) dest += stride; + COPY (dest, vx4) dest += stride; + COPY (dest, vx5) dest += stride; + COPY (dest, vx6) dest += stride; COPY (dest, vx7) POWERPC_PERF_STOP_COUNT(altivec_idct_put_num, 1); @@ -225,22 +225,22 @@ perm0 = vec_mergeh (p, p0); perm1 = vec_mergeh (p, p1); -#define ADD(dest,src,perm) \ - /* *(uint64_t *)&tmp = *(uint64_t *)dest; */ \ - tmp = vec_ld (0, dest); \ - tmp2 = (vector_s16_t)vec_perm (tmp, (vector_u8_t)zero, perm); \ - tmp3 = vec_adds (tmp2, src); \ - tmp = vec_packsu (tmp3, tmp3); \ - vec_ste ((vector_u32_t)tmp, 0, (unsigned int *)dest); \ +#define ADD(dest,src,perm) \ + /* *(uint64_t *)&tmp = *(uint64_t *)dest; */ \ + tmp = vec_ld (0, dest); \ + tmp2 = (vector_s16_t)vec_perm (tmp, (vector_u8_t)zero, perm); \ + tmp3 = vec_adds (tmp2, src); \ + tmp = vec_packsu (tmp3, tmp3); \ + vec_ste ((vector_u32_t)tmp, 0, (unsigned int *)dest); \ vec_ste ((vector_u32_t)tmp, 4, (unsigned int *)dest); - ADD (dest, vx0, perm0) dest += stride; - ADD (dest, vx1, perm1) dest += stride; - ADD (dest, vx2, perm0) dest += stride; - ADD (dest, vx3, perm1) dest += stride; - ADD (dest, vx4, perm0) dest += stride; - ADD (dest, vx5, perm1) dest += stride; - ADD (dest, vx6, perm0) dest += stride; + ADD (dest, vx0, perm0) dest += stride; + ADD (dest, vx1, perm1) dest += stride; + ADD (dest, vx2, perm0) dest += stride; + ADD (dest, vx3, perm1) dest += stride; + ADD (dest, vx4, perm0) dest += stride; + ADD (dest, vx5, perm1) dest += stride; + ADD (dest, vx6, perm0) dest += stride; ADD (dest, vx7, perm1) POWERPC_PERF_STOP_COUNT(altivec_idct_add_num, 1);
--- a/ppc/mpegvideo_altivec.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ppc/mpegvideo_altivec.c Thu Dec 22 01:10:11 2005 +0000 @@ -152,9 +152,9 @@ } // The following block could exist as a separate an altivec dct - // function. However, if we put it inline, the DCT data can remain - // in the vector local variables, as floats, which we'll use during the - // quantize step... + // function. However, if we put it inline, the DCT data can remain + // in the vector local variables, as floats, which we'll use during the + // quantize step... { const vector float vec_0_298631336 = (vector float)FOUROF(0.298631336f); const vector float vec_0_390180644 = (vector float)FOUROF(-0.390180644f); @@ -206,11 +206,11 @@ z1 = vec_madd(vec_add(tmp12, tmp13), vec_0_541196100, (vector float)zero); // dataptr[2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp13, FIX_0_765366865), - // CONST_BITS-PASS1_BITS); + // CONST_BITS-PASS1_BITS); row2 = vec_madd(tmp13, vec_0_765366865, z1); // dataptr[6] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, - FIX_1_847759065), - // CONST_BITS-PASS1_BITS); + // CONST_BITS-PASS1_BITS); row6 = vec_madd(tmp12, vec_1_847759065, z1); z1 = vec_add(tmp4, tmp7); // z1 = tmp4 + tmp7; @@ -315,7 +315,7 @@ } // Load the bias vector (We add 0.5 to the bias so that we're - // rounding when we convert to int, instead of flooring.) + // rounding when we convert to int, instead of flooring.) { vector signed int biasInt; const vector float negOneFloat = (vector float)FOUROF(-1.0f);
--- a/ps2/dsputil_mmi.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ps2/dsputil_mmi.c Thu Dec 22 01:10:11 2005 +0000 @@ -80,7 +80,7 @@ "pextlb $10, $0, $10 \n\t" "sq $10, 80(%1) \n\t" "pextlb $8, $0, $8 \n\t" - "sq $8, 96(%1) \n\t" + "sq $8, 96(%1) \n\t" "pextlb $9, $0, $9 \n\t" "sq $9, 112(%1) \n\t" ".set pop \n\t" @@ -112,7 +112,7 @@ asm volatile ( ".set push \n\t" ".set mips3 \n\t" - "1: \n\t" + "1: \n\t" "ldr $8, 0(%1) \n\t" "add $11, %1, %3 \n\t" "ldl $8, 7(%1) \n\t" @@ -133,7 +133,7 @@ "bgtz %2, 1b \n\t" ".set pop \n\t" : "+r" (block), "+r" (pixels), "+r" (h) : "r" (line_size) - : "$8", "$9", "$10", "$11", "$12", "$13", "memory" ); + : "$8", "$9", "$10", "$11", "$12", "$13", "memory" ); }
--- a/ps2/idct_mmi.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ps2/idct_mmi.c Thu Dec 22 01:10:11 2005 +0000 @@ -15,32 +15,32 @@ #include "../dsputil.h" #include "mmi.h" -#define BITS_INV_ACC 5 // 4 or 5 for IEEE -#define SHIFT_INV_ROW (16 - BITS_INV_ACC) +#define BITS_INV_ACC 5 // 4 or 5 for IEEE +#define SHIFT_INV_ROW (16 - BITS_INV_ACC) #define SHIFT_INV_COL (1 + BITS_INV_ACC) -#define TG1 6518 -#define TG2 13573 -#define TG3 21895 -#define CS4 23170 +#define TG1 6518 +#define TG2 13573 +#define TG3 21895 +#define CS4 23170 -#define ROUNDER_0 0 -#define ROUNDER_1 16 +#define ROUNDER_0 0 +#define ROUNDER_1 16 -#define TAB_i_04 (32+0) -#define TAB_i_17 (32+64) -#define TAB_i_26 (32+128) -#define TAB_i_35 (32+192) +#define TAB_i_04 (32+0) +#define TAB_i_17 (32+64) +#define TAB_i_26 (32+128) +#define TAB_i_35 (32+192) -#define TG_1_16 (32+256+0) -#define TG_2_16 (32+256+16) -#define TG_3_16 (32+256+32) -#define COS_4_16 (32+256+48) +#define TG_1_16 (32+256+0) +#define TG_2_16 (32+256+16) +#define TG_3_16 (32+256+32) +#define COS_4_16 (32+256+48) -#define CLIPMAX (32+256+64+0) +#define CLIPMAX (32+256+64+0) static short consttable[] align16 = { -/* rounder 0*/ // assume SHIFT_INV_ROW == 11 +/* rounder 0*/ // assume SHIFT_INV_ROW == 11 0x3ff, 1, 0x3ff, 1, 0x3ff, 1, 0x3ff, 1, /* rounder 1*/ 0x3ff, 0, 0x3ff, 0, 0x3ff, 0, 0x3ff, 0, @@ -75,274 +75,274 @@ #define DCT_8_INV_ROW1(blk, rowoff, taboff, rnd, outreg) { \ - lq(blk, rowoff, $16); /* r16 = x7 x5 x3 x1 x6 x4 x2 x0 */ \ - /*slot*/ \ - lq($24, 0+taboff, $17); /* r17 = w */ \ - /*delay slot $16*/ \ - lq($24, 16+taboff, $18);/* r18 = w */ \ - prevh($16, $2); /* r2 = x1 x3 x5 x7 x0 x2 x4 x6 */ \ - lq($24, 32+taboff, $19);/* r19 = w */ \ - phmadh($17, $16, $17); /* r17 = b1"b0'a1"a0' */ \ - lq($24, 48+taboff, $20);/* r20 = w */ \ - phmadh($18, $2, $18); /* r18 = b1'b0"a1'a0" */ \ - phmadh($19, $16, $19); /* r19 = b3"b2'a3"a2' */ \ - phmadh($20, $2, $20); /* r20 = b3'b2"a3'a2" */ \ - paddw($17, $18, $17); /* r17 = (b1)(b0)(a1)(a0) */ \ - paddw($19, $20, $19); /* r19 = (b3)(b2)(a3)(a2) */ \ - pcpyld($19, $17, $18); /* r18 = (a3)(a2)(a1)(a0) */ \ - pcpyud($17, $19, $20); /* r20 = (b3)(b2)(b1)(b0) */ \ - paddw($18, rnd, $18); /* r18 = (a3)(a2)(a1)(a0) */\ - paddw($18, $20, $17); /* r17 = ()()()(a0+b0) */ \ - psubw($18, $20, $20); /* r20 = ()()()(a0-b0) */ \ - psraw($17, SHIFT_INV_ROW, $17); /* r17 = (y3 y2 y1 y0) */ \ - psraw($20, SHIFT_INV_ROW, $20); /* r20 = (y4 y5 y6 y7) */ \ - ppach($20, $17, outreg);/* out = y4 y5 y6 y7 y3 y2 y1 y0 Note order */ \ + lq(blk, rowoff, $16); /* r16 = x7 x5 x3 x1 x6 x4 x2 x0 */ \ + /*slot*/ \ + lq($24, 0+taboff, $17); /* r17 = w */ \ + /*delay slot $16*/ \ + lq($24, 16+taboff, $18);/* r18 = w */ \ + prevh($16, $2); /* r2 = x1 x3 x5 x7 x0 x2 x4 x6 */ \ + lq($24, 32+taboff, $19);/* r19 = w */ \ + phmadh($17, $16, $17); /* r17 = b1"b0'a1"a0' */ \ + lq($24, 48+taboff, $20);/* r20 = w */ \ + phmadh($18, $2, $18); /* r18 = b1'b0"a1'a0" */ \ + phmadh($19, $16, $19); /* r19 = b3"b2'a3"a2' */ \ + phmadh($20, $2, $20); /* r20 = b3'b2"a3'a2" */ \ + paddw($17, $18, $17); /* r17 = (b1)(b0)(a1)(a0) */ \ + paddw($19, $20, $19); /* r19 = (b3)(b2)(a3)(a2) */ \ + pcpyld($19, $17, $18); /* r18 = (a3)(a2)(a1)(a0) */ \ + pcpyud($17, $19, $20); /* r20 = (b3)(b2)(b1)(b0) */ \ + paddw($18, rnd, $18); /* r18 = (a3)(a2)(a1)(a0) */\ + paddw($18, $20, $17); /* r17 = ()()()(a0+b0) */ \ + psubw($18, $20, $20); /* r20 = ()()()(a0-b0) */ \ + psraw($17, SHIFT_INV_ROW, $17); /* r17 = (y3 y2 y1 y0) */ \ + psraw($20, SHIFT_INV_ROW, $20); /* r20 = (y4 y5 y6 y7) */ \ + ppach($20, $17, outreg);/* out = y4 y5 y6 y7 y3 y2 y1 y0 Note order */ \ \ - prevh(outreg, $2); \ - pcpyud($2, $2, $2); \ - pcpyld($2, outreg, outreg); \ + prevh(outreg, $2); \ + pcpyud($2, $2, $2); \ + pcpyld($2, outreg, outreg); \ } #define DCT_8_INV_COL8() \ \ - lq($24, TG_3_16, $2); /* r2 = tn3 */ \ + lq($24, TG_3_16, $2); /* r2 = tn3 */ \ \ - pmulth($11, $2, $17); /* r17 = x3 * tn3 (6420) */ \ - psraw($17, 15, $17); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $17, $17); /* r17 = x3 * tn3 */ \ - psubh($17, $13, $17); /* r17 = tm35 */ \ + pmulth($11, $2, $17); /* r17 = x3 * tn3 (6420) */ \ + psraw($17, 15, $17); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $17, $17); /* r17 = x3 * tn3 */ \ + psubh($17, $13, $17); /* r17 = tm35 */ \ \ - pmulth($13, $2, $18); /* r18 = x5 * tn3 (6420) */ \ - psraw($18, 15, $18); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $18, $18); /* r18 = x5 * tn3 */ \ - paddh($18, $11, $18); /* r18 = tp35 */ \ + pmulth($13, $2, $18); /* r18 = x5 * tn3 (6420) */ \ + psraw($18, 15, $18); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $18, $18); /* r18 = x5 * tn3 */ \ + paddh($18, $11, $18); /* r18 = tp35 */ \ \ - lq($24, TG_1_16, $2); /* r2 = tn1 */ \ + lq($24, TG_1_16, $2); /* r2 = tn1 */ \ \ - pmulth($15, $2, $19); /* r19 = x7 * tn1 (6420) */ \ - psraw($19, 15, $19); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $19, $19); /* r19 = x7 * tn1 */ \ - paddh($19, $9, $19); /* r19 = tp17 */ \ + pmulth($15, $2, $19); /* r19 = x7 * tn1 (6420) */ \ + psraw($19, 15, $19); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $19, $19); /* r19 = x7 * tn1 */ \ + paddh($19, $9, $19); /* r19 = tp17 */ \ \ - pmulth($9, $2, $20); /* r20 = x1 * tn1 (6420) */ \ - psraw($20, 15, $20); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $20, $20); /* r20 = x1 * tn1 */ \ - psubh($20, $15, $20); /* r20 = tm17 */ \ + pmulth($9, $2, $20); /* r20 = x1 * tn1 (6420) */ \ + psraw($20, 15, $20); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $20, $20); /* r20 = x1 * tn1 */ \ + psubh($20, $15, $20); /* r20 = tm17 */ \ \ - psubh($19, $18, $3); /* r3 = t1 */ \ - paddh($20, $17, $16); /* r16 = t2 */ \ - psubh($20, $17, $23); /* r23 = b3 */ \ - paddh($19, $18, $20); /* r20 = b0 */ \ + psubh($19, $18, $3); /* r3 = t1 */ \ + paddh($20, $17, $16); /* r16 = t2 */ \ + psubh($20, $17, $23); /* r23 = b3 */ \ + paddh($19, $18, $20); /* r20 = b0 */ \ \ - lq($24, COS_4_16, $2); /* r2 = cs4 */ \ + lq($24, COS_4_16, $2); /* r2 = cs4 */ \ \ - paddh($3, $16, $21); /* r21 = t1+t2 */ \ - psubh($3, $16, $22); /* r22 = t1-t2 */ \ + paddh($3, $16, $21); /* r21 = t1+t2 */ \ + psubh($3, $16, $22); /* r22 = t1-t2 */ \ \ - pmulth($21, $2, $21); /* r21 = cs4 * (t1+t2) 6420 */ \ - psraw($21, 15, $21); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $21, $21); /* r21 = b1 */ \ + pmulth($21, $2, $21); /* r21 = cs4 * (t1+t2) 6420 */ \ + psraw($21, 15, $21); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $21, $21); /* r21 = b1 */ \ \ - pmulth($22, $2, $22); /* r22 = cs4 * (t1-t2) 6420 */ \ - psraw($22, 15, $22); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $22, $22); /* r22 = b2 */ \ + pmulth($22, $2, $22); /* r22 = cs4 * (t1-t2) 6420 */ \ + psraw($22, 15, $22); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $22, $22); /* r22 = b2 */ \ \ - lq($24, TG_2_16, $2); /* r2 = tn2 */ \ + lq($24, TG_2_16, $2); /* r2 = tn2 */ \ \ - pmulth($10, $2, $17); /* r17 = x2 * tn2 (6420) */ \ - psraw($17, 15, $17); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $17, $17); /* r17 = x3 * tn3 */ \ - psubh($17, $14, $17); /* r17 = tm26 */ \ + pmulth($10, $2, $17); /* r17 = x2 * tn2 (6420) */ \ + psraw($17, 15, $17); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $17, $17); /* r17 = x3 * tn3 */ \ + psubh($17, $14, $17); /* r17 = tm26 */ \ \ - pmulth($14, $2, $18); /* r18 = x6 * tn2 (6420) */ \ - psraw($18, 15, $18); \ - pmfhl_uw($3); /* r3 = 7531 */ \ - psraw($3, 15, $3); \ - pinteh($3, $18, $18); /* r18 = x6 * tn2 */ \ - paddh($18, $10, $18); /* r18 = tp26 */ \ + pmulth($14, $2, $18); /* r18 = x6 * tn2 (6420) */ \ + psraw($18, 15, $18); \ + pmfhl_uw($3); /* r3 = 7531 */ \ + psraw($3, 15, $3); \ + pinteh($3, $18, $18); /* r18 = x6 * tn2 */ \ + paddh($18, $10, $18); /* r18 = tp26 */ \ \ - paddh($8, $12, $2); /* r2 = tp04 */ \ - psubh($8, $12, $3); /* r3 = tm04 */ \ + paddh($8, $12, $2); /* r2 = tp04 */ \ + psubh($8, $12, $3); /* r3 = tm04 */ \ \ - paddh($2, $18, $16); /* r16 = a0 */ \ - psubh($2, $18, $19); /* r19 = a3 */ \ - psubh($3, $17, $18); /* r18 = a2 */ \ - paddh($3, $17, $17); /* r17 = a1 */ + paddh($2, $18, $16); /* r16 = a0 */ \ + psubh($2, $18, $19); /* r19 = a3 */ \ + psubh($3, $17, $18); /* r18 = a2 */ \ + paddh($3, $17, $17); /* r17 = a1 */ #define DCT_8_INV_COL8_STORE(blk) \ \ - paddh($16, $20, $2); /* y0 a0+b0 */ \ - psubh($16, $20, $16); /* y7 a0-b0 */ \ - psrah($2, SHIFT_INV_COL, $2); \ - psrah($16, SHIFT_INV_COL, $16); \ - sq($2, 0, blk); \ - sq($16, 112, blk); \ + paddh($16, $20, $2); /* y0 a0+b0 */ \ + psubh($16, $20, $16); /* y7 a0-b0 */ \ + psrah($2, SHIFT_INV_COL, $2); \ + psrah($16, SHIFT_INV_COL, $16); \ + sq($2, 0, blk); \ + sq($16, 112, blk); \ \ - paddh($17, $21, $3); /* y1 a1+b1 */ \ - psubh($17, $21, $17); /* y6 a1-b1 */ \ - psrah($3, SHIFT_INV_COL, $3); \ - psrah($17, SHIFT_INV_COL, $17); \ - sq($3, 16, blk); \ - sq($17, 96, blk); \ + paddh($17, $21, $3); /* y1 a1+b1 */ \ + psubh($17, $21, $17); /* y6 a1-b1 */ \ + psrah($3, SHIFT_INV_COL, $3); \ + psrah($17, SHIFT_INV_COL, $17); \ + sq($3, 16, blk); \ + sq($17, 96, blk); \ \ - paddh($18, $22, $2); /* y2 a2+b2 */ \ - psubh($18, $22, $18); /* y5 a2-b2 */ \ - psrah($2, SHIFT_INV_COL, $2); \ - psrah($18, SHIFT_INV_COL, $18); \ - sq($2, 32, blk); \ - sq($18, 80, blk); \ + paddh($18, $22, $2); /* y2 a2+b2 */ \ + psubh($18, $22, $18); /* y5 a2-b2 */ \ + psrah($2, SHIFT_INV_COL, $2); \ + psrah($18, SHIFT_INV_COL, $18); \ + sq($2, 32, blk); \ + sq($18, 80, blk); \ \ - paddh($19, $23, $3); /* y3 a3+b3 */ \ - psubh($19, $23, $19); /* y4 a3-b3 */ \ - psrah($3, SHIFT_INV_COL, $3); \ - psrah($19, SHIFT_INV_COL, $19); \ - sq($3, 48, blk); \ - sq($19, 64, blk); + paddh($19, $23, $3); /* y3 a3+b3 */ \ + psubh($19, $23, $19); /* y4 a3-b3 */ \ + psrah($3, SHIFT_INV_COL, $3); \ + psrah($19, SHIFT_INV_COL, $19); \ + sq($3, 48, blk); \ + sq($19, 64, blk); #define DCT_8_INV_COL8_PMS() \ - paddh($16, $20, $2); /* y0 a0+b0 */ \ - psubh($16, $20, $20); /* y7 a0-b0 */ \ - psrah($2, SHIFT_INV_COL, $16); \ - psrah($20, SHIFT_INV_COL, $20); \ + paddh($16, $20, $2); /* y0 a0+b0 */ \ + psubh($16, $20, $20); /* y7 a0-b0 */ \ + psrah($2, SHIFT_INV_COL, $16); \ + psrah($20, SHIFT_INV_COL, $20); \ \ - paddh($17, $21, $3); /* y1 a1+b1 */ \ - psubh($17, $21, $21); /* y6 a1-b1 */ \ - psrah($3, SHIFT_INV_COL, $17); \ - psrah($21, SHIFT_INV_COL, $21); \ + paddh($17, $21, $3); /* y1 a1+b1 */ \ + psubh($17, $21, $21); /* y6 a1-b1 */ \ + psrah($3, SHIFT_INV_COL, $17); \ + psrah($21, SHIFT_INV_COL, $21); \ \ - paddh($18, $22, $2); /* y2 a2+b2 */ \ - psubh($18, $22, $22); /* y5 a2-b2 */ \ - psrah($2, SHIFT_INV_COL, $18); \ - psrah($22, SHIFT_INV_COL, $22); \ + paddh($18, $22, $2); /* y2 a2+b2 */ \ + psubh($18, $22, $22); /* y5 a2-b2 */ \ + psrah($2, SHIFT_INV_COL, $18); \ + psrah($22, SHIFT_INV_COL, $22); \ \ - paddh($19, $23, $3); /* y3 a3+b3 */ \ - psubh($19, $23, $23); /* y4 a3-b3 */ \ - psrah($3, SHIFT_INV_COL, $19); \ - psrah($23, SHIFT_INV_COL, $23); + paddh($19, $23, $3); /* y3 a3+b3 */ \ + psubh($19, $23, $23); /* y4 a3-b3 */ \ + psrah($3, SHIFT_INV_COL, $19); \ + psrah($23, SHIFT_INV_COL, $23); -#define PUT(rs) \ - pminh(rs, $11, $2); \ - pmaxh($2, $0, $2); \ - ppacb($0, $2, $2); \ - sd3(2, 0, 4); \ - __asm__ __volatile__ ("add $4, $5, $4"); +#define PUT(rs) \ + pminh(rs, $11, $2); \ + pmaxh($2, $0, $2); \ + ppacb($0, $2, $2); \ + sd3(2, 0, 4); \ + __asm__ __volatile__ ("add $4, $5, $4"); #define DCT_8_INV_COL8_PUT() \ - PUT($16); \ - PUT($17); \ - PUT($18); \ - PUT($19); \ - PUT($23); \ - PUT($22); \ - PUT($21); \ - PUT($20); + PUT($16); \ + PUT($17); \ + PUT($18); \ + PUT($19); \ + PUT($23); \ + PUT($22); \ + PUT($21); \ + PUT($20); -#define ADD(rs) \ - ld3(4, 0, 2); \ - pextlb($0, $2, $2); \ - paddh($2, rs, $2); \ - pminh($2, $11, $2); \ - pmaxh($2, $0, $2); \ - ppacb($0, $2, $2); \ - sd3(2, 0, 4); \ - __asm__ __volatile__ ("add $4, $5, $4"); +#define ADD(rs) \ + ld3(4, 0, 2); \ + pextlb($0, $2, $2); \ + paddh($2, rs, $2); \ + pminh($2, $11, $2); \ + pmaxh($2, $0, $2); \ + ppacb($0, $2, $2); \ + sd3(2, 0, 4); \ + __asm__ __volatile__ ("add $4, $5, $4"); /*fixme: schedule*/ #define DCT_8_INV_COL8_ADD() \ - ADD($16); \ - ADD($17); \ - ADD($18); \ - ADD($19); \ - ADD($23); \ - ADD($22); \ - ADD($21); \ - ADD($20); + ADD($16); \ + ADD($17); \ + ADD($18); \ + ADD($19); \ + ADD($23); \ + ADD($22); \ + ADD($21); \ + ADD($20); void ff_mmi_idct(int16_t * block) { - /* $4 = block */ - __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); - lq($24, ROUNDER_0, $8); - lq($24, ROUNDER_1, $7); - DCT_8_INV_ROW1($4, 0, TAB_i_04, $8, $8); - DCT_8_INV_ROW1($4, 16, TAB_i_17, $7, $9); - DCT_8_INV_ROW1($4, 32, TAB_i_26, $7, $10); - DCT_8_INV_ROW1($4, 48, TAB_i_35, $7, $11); - DCT_8_INV_ROW1($4, 64, TAB_i_04, $7, $12); - DCT_8_INV_ROW1($4, 80, TAB_i_35, $7, $13); - DCT_8_INV_ROW1($4, 96, TAB_i_26, $7, $14); - DCT_8_INV_ROW1($4, 112, TAB_i_17, $7, $15); - DCT_8_INV_COL8(); - DCT_8_INV_COL8_STORE($4); + /* $4 = block */ + __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); + lq($24, ROUNDER_0, $8); + lq($24, ROUNDER_1, $7); + DCT_8_INV_ROW1($4, 0, TAB_i_04, $8, $8); + DCT_8_INV_ROW1($4, 16, TAB_i_17, $7, $9); + DCT_8_INV_ROW1($4, 32, TAB_i_26, $7, $10); + DCT_8_INV_ROW1($4, 48, TAB_i_35, $7, $11); + DCT_8_INV_ROW1($4, 64, TAB_i_04, $7, $12); + DCT_8_INV_ROW1($4, 80, TAB_i_35, $7, $13); + DCT_8_INV_ROW1($4, 96, TAB_i_26, $7, $14); + DCT_8_INV_ROW1($4, 112, TAB_i_17, $7, $15); + DCT_8_INV_COL8(); + DCT_8_INV_COL8_STORE($4); - //let savedtemp regs be saved - __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); + //let savedtemp regs be saved + __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); } void ff_mmi_idct_put(uint8_t *dest, int line_size, DCTELEM *block) { - /* $4 = dest, $5 = line_size, $6 = block */ - __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); - lq($24, ROUNDER_0, $8); - lq($24, ROUNDER_1, $7); - DCT_8_INV_ROW1($6, 0, TAB_i_04, $8, $8); - DCT_8_INV_ROW1($6, 16, TAB_i_17, $7, $9); - DCT_8_INV_ROW1($6, 32, TAB_i_26, $7, $10); - DCT_8_INV_ROW1($6, 48, TAB_i_35, $7, $11); - DCT_8_INV_ROW1($6, 64, TAB_i_04, $7, $12); - DCT_8_INV_ROW1($6, 80, TAB_i_35, $7, $13); - DCT_8_INV_ROW1($6, 96, TAB_i_26, $7, $14); - DCT_8_INV_ROW1($6, 112, TAB_i_17, $7, $15); - DCT_8_INV_COL8(); - lq($24, CLIPMAX, $11); - DCT_8_INV_COL8_PMS(); - DCT_8_INV_COL8_PUT(); + /* $4 = dest, $5 = line_size, $6 = block */ + __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); + lq($24, ROUNDER_0, $8); + lq($24, ROUNDER_1, $7); + DCT_8_INV_ROW1($6, 0, TAB_i_04, $8, $8); + DCT_8_INV_ROW1($6, 16, TAB_i_17, $7, $9); + DCT_8_INV_ROW1($6, 32, TAB_i_26, $7, $10); + DCT_8_INV_ROW1($6, 48, TAB_i_35, $7, $11); + DCT_8_INV_ROW1($6, 64, TAB_i_04, $7, $12); + DCT_8_INV_ROW1($6, 80, TAB_i_35, $7, $13); + DCT_8_INV_ROW1($6, 96, TAB_i_26, $7, $14); + DCT_8_INV_ROW1($6, 112, TAB_i_17, $7, $15); + DCT_8_INV_COL8(); + lq($24, CLIPMAX, $11); + DCT_8_INV_COL8_PMS(); + DCT_8_INV_COL8_PUT(); - //let savedtemp regs be saved - __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); + //let savedtemp regs be saved + __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); } void ff_mmi_idct_add(uint8_t *dest, int line_size, DCTELEM *block) { - /* $4 = dest, $5 = line_size, $6 = block */ - __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); - lq($24, ROUNDER_0, $8); - lq($24, ROUNDER_1, $7); - DCT_8_INV_ROW1($6, 0, TAB_i_04, $8, $8); - DCT_8_INV_ROW1($6, 16, TAB_i_17, $7, $9); - DCT_8_INV_ROW1($6, 32, TAB_i_26, $7, $10); - DCT_8_INV_ROW1($6, 48, TAB_i_35, $7, $11); - DCT_8_INV_ROW1($6, 64, TAB_i_04, $7, $12); - DCT_8_INV_ROW1($6, 80, TAB_i_35, $7, $13); - DCT_8_INV_ROW1($6, 96, TAB_i_26, $7, $14); - DCT_8_INV_ROW1($6, 112, TAB_i_17, $7, $15); - DCT_8_INV_COL8(); - lq($24, CLIPMAX, $11); - DCT_8_INV_COL8_PMS(); - DCT_8_INV_COL8_ADD(); + /* $4 = dest, $5 = line_size, $6 = block */ + __asm__ __volatile__("la $24, %0"::"m"(consttable[0])); + lq($24, ROUNDER_0, $8); + lq($24, ROUNDER_1, $7); + DCT_8_INV_ROW1($6, 0, TAB_i_04, $8, $8); + DCT_8_INV_ROW1($6, 16, TAB_i_17, $7, $9); + DCT_8_INV_ROW1($6, 32, TAB_i_26, $7, $10); + DCT_8_INV_ROW1($6, 48, TAB_i_35, $7, $11); + DCT_8_INV_ROW1($6, 64, TAB_i_04, $7, $12); + DCT_8_INV_ROW1($6, 80, TAB_i_35, $7, $13); + DCT_8_INV_ROW1($6, 96, TAB_i_26, $7, $14); + DCT_8_INV_ROW1($6, 112, TAB_i_17, $7, $15); + DCT_8_INV_COL8(); + lq($24, CLIPMAX, $11); + DCT_8_INV_COL8_PMS(); + DCT_8_INV_COL8_ADD(); - //let savedtemp regs be saved - __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); + //let savedtemp regs be saved + __asm__ __volatile__(" ":::"$16", "$17", "$18", "$19", "$20", "$21", "$22", "$23"); }
--- a/ps2/mmi.h Wed Dec 21 17:50:40 2005 +0000 +++ b/ps2/mmi.h Thu Dec 22 01:10:11 2005 +0000 @@ -5,148 +5,148 @@ /* #define r0 $zero -#define r1 $at //assembler! -#define r2 $v0 //return -#define r3 $v1 //return -#define r4 $a0 //arg -#define r5 $a1 //arg -#define r6 $a2 //arg -#define r7 $a3 //arg -#define r8 $t0 //temp -#define r9 $t1 //temp -#define r10 $t2 //temp -#define r11 $t3 //temp -#define r12 $t4 //temp -#define r13 $t5 //temp -#define r14 $t6 //temp -#define r15 $t7 //temp -#define r16 $s0 //saved temp -#define r17 $s1 //saved temp -#define r18 $s2 //saved temp -#define r19 $s3 //saved temp -#define r20 $s4 //saved temp -#define r21 $s5 //saved temp -#define r22 $s6 //saved temp -#define r23 $s7 //saved temp -#define r24 $t8 //temp -#define r25 $t9 //temp -#define r26 $k0 //kernel -#define r27 $k1 //kernel -#define r28 $gp //global ptr -#define r29 $sp //stack ptr -#define r30 $fp //frame ptr -#define r31 $ra //return addr +#define r1 $at //assembler! +#define r2 $v0 //return +#define r3 $v1 //return +#define r4 $a0 //arg +#define r5 $a1 //arg +#define r6 $a2 //arg +#define r7 $a3 //arg +#define r8 $t0 //temp +#define r9 $t1 //temp +#define r10 $t2 //temp +#define r11 $t3 //temp +#define r12 $t4 //temp +#define r13 $t5 //temp +#define r14 $t6 //temp +#define r15 $t7 //temp +#define r16 $s0 //saved temp +#define r17 $s1 //saved temp +#define r18 $s2 //saved temp +#define r19 $s3 //saved temp +#define r20 $s4 //saved temp +#define r21 $s5 //saved temp +#define r22 $s6 //saved temp +#define r23 $s7 //saved temp +#define r24 $t8 //temp +#define r25 $t9 //temp +#define r26 $k0 //kernel +#define r27 $k1 //kernel +#define r28 $gp //global ptr +#define r29 $sp //stack ptr +#define r30 $fp //frame ptr +#define r31 $ra //return addr */ -#define lq(base, off, reg) \ - __asm__ __volatile__ ("lq " #reg ", %0("#base ")" : : "i" (off) ) +#define lq(base, off, reg) \ + __asm__ __volatile__ ("lq " #reg ", %0("#base ")" : : "i" (off) ) -#define lq2(mem, reg) \ - __asm__ __volatile__ ("lq " #reg ", %0" : : "r" (mem)) +#define lq2(mem, reg) \ + __asm__ __volatile__ ("lq " #reg ", %0" : : "r" (mem)) -#define sq(reg, off, base) \ - __asm__ __volatile__ ("sq " #reg ", %0("#base ")" : : "i" (off) ) +#define sq(reg, off, base) \ + __asm__ __volatile__ ("sq " #reg ", %0("#base ")" : : "i" (off) ) /* -#define ld(base, off, reg) \ - __asm__ __volatile__ ("ld " #reg ", " #off "("#base ")") +#define ld(base, off, reg) \ + __asm__ __volatile__ ("ld " #reg ", " #off "("#base ")") */ -#define ld3(base, off, reg) \ - __asm__ __volatile__ (".word %0" : : "i" ( 0xdc000000 | (base<<21) | (reg<<16) | (off))) +#define ld3(base, off, reg) \ + __asm__ __volatile__ (".word %0" : : "i" ( 0xdc000000 | (base<<21) | (reg<<16) | (off))) -#define ldr3(base, off, reg) \ - __asm__ __volatile__ (".word %0" : : "i" ( 0x6c000000 | (base<<21) | (reg<<16) | (off))) +#define ldr3(base, off, reg) \ + __asm__ __volatile__ (".word %0" : : "i" ( 0x6c000000 | (base<<21) | (reg<<16) | (off))) -#define ldl3(base, off, reg) \ - __asm__ __volatile__ (".word %0" : : "i" ( 0x68000000 | (base<<21) | (reg<<16) | (off))) +#define ldl3(base, off, reg) \ + __asm__ __volatile__ (".word %0" : : "i" ( 0x68000000 | (base<<21) | (reg<<16) | (off))) /* -#define sd(reg, off, base) \ - __asm__ __volatile__ ("sd " #reg ", " #off "("#base ")") +#define sd(reg, off, base) \ + __asm__ __volatile__ ("sd " #reg ", " #off "("#base ")") */ //seems assembler has bug encoding mnemonic 'sd', so DIY -#define sd3(reg, off, base) \ - __asm__ __volatile__ (".word %0" : : "i" ( 0xfc000000 | (base<<21) | (reg<<16) | (off))) +#define sd3(reg, off, base) \ + __asm__ __volatile__ (".word %0" : : "i" ( 0xfc000000 | (base<<21) | (reg<<16) | (off))) -#define sw(reg, off, base) \ - __asm__ __volatile__ ("sw " #reg ", " #off "("#base ")") +#define sw(reg, off, base) \ + __asm__ __volatile__ ("sw " #reg ", " #off "("#base ")") -#define sq2(reg, mem) \ - __asm__ __volatile__ ("sq " #reg ", %0" : : "m" (*(mem))) +#define sq2(reg, mem) \ + __asm__ __volatile__ ("sq " #reg ", %0" : : "m" (*(mem))) -#define pinth(rs, rt, rd) \ - __asm__ __volatile__ ("pinth " #rd ", " #rs ", " #rt ) +#define pinth(rs, rt, rd) \ + __asm__ __volatile__ ("pinth " #rd ", " #rs ", " #rt ) -#define phmadh(rs, rt, rd) \ - __asm__ __volatile__ ("phmadh " #rd ", " #rs ", " #rt ) +#define phmadh(rs, rt, rd) \ + __asm__ __volatile__ ("phmadh " #rd ", " #rs ", " #rt ) -#define pcpyud(rs, rt, rd) \ - __asm__ __volatile__ ("pcpyud " #rd ", " #rs ", " #rt ) +#define pcpyud(rs, rt, rd) \ + __asm__ __volatile__ ("pcpyud " #rd ", " #rs ", " #rt ) -#define pcpyld(rs, rt, rd) \ - __asm__ __volatile__ ("pcpyld " #rd ", " #rs ", " #rt ) +#define pcpyld(rs, rt, rd) \ + __asm__ __volatile__ ("pcpyld " #rd ", " #rs ", " #rt ) -#define pcpyh(rt, rd) \ - __asm__ __volatile__ ("pcpyh " #rd ", " #rt ) +#define pcpyh(rt, rd) \ + __asm__ __volatile__ ("pcpyh " #rd ", " #rt ) -#define paddw(rs, rt, rd) \ - __asm__ __volatile__ ("paddw " #rd ", " #rs ", " #rt ) +#define paddw(rs, rt, rd) \ + __asm__ __volatile__ ("paddw " #rd ", " #rs ", " #rt ) -#define pextlw(rs, rt, rd) \ - __asm__ __volatile__ ("pextlw " #rd ", " #rs ", " #rt ) +#define pextlw(rs, rt, rd) \ + __asm__ __volatile__ ("pextlw " #rd ", " #rs ", " #rt ) -#define pextuw(rs, rt, rd) \ - __asm__ __volatile__ ("pextuw " #rd ", " #rs ", " #rt ) +#define pextuw(rs, rt, rd) \ + __asm__ __volatile__ ("pextuw " #rd ", " #rs ", " #rt ) -#define pextlh(rs, rt, rd) \ - __asm__ __volatile__ ("pextlh " #rd ", " #rs ", " #rt ) +#define pextlh(rs, rt, rd) \ + __asm__ __volatile__ ("pextlh " #rd ", " #rs ", " #rt ) -#define pextuh(rs, rt, rd) \ - __asm__ __volatile__ ("pextuh " #rd ", " #rs ", " #rt ) +#define pextuh(rs, rt, rd) \ + __asm__ __volatile__ ("pextuh " #rd ", " #rs ", " #rt ) -#define psubw(rs, rt, rd) \ - __asm__ __volatile__ ("psubw " #rd ", " #rs ", " #rt ) +#define psubw(rs, rt, rd) \ + __asm__ __volatile__ ("psubw " #rd ", " #rs ", " #rt ) -#define psraw(rt, sa, rd) \ - __asm__ __volatile__ ("psraw " #rd ", " #rt ", %0" : : "i"(sa) ) +#define psraw(rt, sa, rd) \ + __asm__ __volatile__ ("psraw " #rd ", " #rt ", %0" : : "i"(sa) ) -#define ppach(rs, rt, rd) \ - __asm__ __volatile__ ("ppach " #rd ", " #rs ", " #rt ) +#define ppach(rs, rt, rd) \ + __asm__ __volatile__ ("ppach " #rd ", " #rs ", " #rt ) -#define ppacb(rs, rt, rd) \ - __asm__ __volatile__ ("ppacb " #rd ", " #rs ", " #rt ) +#define ppacb(rs, rt, rd) \ + __asm__ __volatile__ ("ppacb " #rd ", " #rs ", " #rt ) -#define prevh(rt, rd) \ - __asm__ __volatile__ ("prevh " #rd ", " #rt ) +#define prevh(rt, rd) \ + __asm__ __volatile__ ("prevh " #rd ", " #rt ) -#define pmulth(rs, rt, rd) \ - __asm__ __volatile__ ("pmulth " #rd ", " #rs ", " #rt ) +#define pmulth(rs, rt, rd) \ + __asm__ __volatile__ ("pmulth " #rd ", " #rs ", " #rt ) -#define pmaxh(rs, rt, rd) \ - __asm__ __volatile__ ("pmaxh " #rd ", " #rs ", " #rt ) +#define pmaxh(rs, rt, rd) \ + __asm__ __volatile__ ("pmaxh " #rd ", " #rs ", " #rt ) -#define pminh(rs, rt, rd) \ - __asm__ __volatile__ ("pminh " #rd ", " #rs ", " #rt ) +#define pminh(rs, rt, rd) \ + __asm__ __volatile__ ("pminh " #rd ", " #rs ", " #rt ) -#define pinteh(rs, rt, rd) \ - __asm__ __volatile__ ("pinteh " #rd ", " #rs ", " #rt ) +#define pinteh(rs, rt, rd) \ + __asm__ __volatile__ ("pinteh " #rd ", " #rs ", " #rt ) -#define paddh(rs, rt, rd) \ - __asm__ __volatile__ ("paddh " #rd ", " #rs ", " #rt ) +#define paddh(rs, rt, rd) \ + __asm__ __volatile__ ("paddh " #rd ", " #rs ", " #rt ) -#define psubh(rs, rt, rd) \ - __asm__ __volatile__ ("psubh " #rd ", " #rs ", " #rt ) +#define psubh(rs, rt, rd) \ + __asm__ __volatile__ ("psubh " #rd ", " #rs ", " #rt ) -#define psrah(rt, sa, rd) \ - __asm__ __volatile__ ("psrah " #rd ", " #rt ", %0" : : "i"(sa) ) +#define psrah(rt, sa, rd) \ + __asm__ __volatile__ ("psrah " #rd ", " #rt ", %0" : : "i"(sa) ) -#define pmfhl_uw(rd) \ - __asm__ __volatile__ ("pmfhl.uw " #rd) +#define pmfhl_uw(rd) \ + __asm__ __volatile__ ("pmfhl.uw " #rd) -#define pextlb(rs, rt, rd) \ - __asm__ __volatile__ ("pextlb " #rd ", " #rs ", " #rt ) +#define pextlb(rs, rt, rd) \ + __asm__ __volatile__ ("pextlb " #rd ", " #rs ", " #rt ) #endif
--- a/ps2/mpegvideo_mmi.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ps2/mpegvideo_mmi.c Thu Dec 22 01:10:11 2005 +0000 @@ -41,7 +41,7 @@ level = block[0] * s->c_dc_scale; }else { qadd = 0; - level = block[0]; + level = block[0]; } nCoeffs= 63; //does not allways use zigzag table } else { @@ -49,29 +49,29 @@ } asm volatile( - "add $14, $0, %3 \n\t" - "pcpyld $8, %0, %0 \n\t" - "pcpyh $8, $8 \n\t" //r8 = qmul - "pcpyld $9, %1, %1 \n\t" - "pcpyh $9, $9 \n\t" //r9 = qadd + "add $14, $0, %3 \n\t" + "pcpyld $8, %0, %0 \n\t" + "pcpyh $8, $8 \n\t" //r8 = qmul + "pcpyld $9, %1, %1 \n\t" + "pcpyh $9, $9 \n\t" //r9 = qadd ".p2align 2 \n\t" - "1: \n\t" - "lq $10, 0($14) \n\t" //r10 = level - "addi $14, $14, 16 \n\t" //block+=8 - "addi %2, %2, -8 \n\t" - "pcgth $11, $0, $10 \n\t" //r11 = level < 0 ? -1 : 0 - "pcgth $12, $10, $0 \n\t" //r12 = level > 0 ? -1 : 0 - "por $12, $11, $12 \n\t" - "pmulth $10, $10, $8 \n\t" - "paddh $13, $9, $11 \n\t" + "1: \n\t" + "lq $10, 0($14) \n\t" //r10 = level + "addi $14, $14, 16 \n\t" //block+=8 + "addi %2, %2, -8 \n\t" + "pcgth $11, $0, $10 \n\t" //r11 = level < 0 ? -1 : 0 + "pcgth $12, $10, $0 \n\t" //r12 = level > 0 ? -1 : 0 + "por $12, $11, $12 \n\t" + "pmulth $10, $10, $8 \n\t" + "paddh $13, $9, $11 \n\t" "pxor $13, $13, $11 \n\t" //r13 = level < 0 ? -qadd : qadd - "pmfhl.uw $11 \n\t" - "pinteh $10, $11, $10 \n\t" //r10 = level * qmul - "paddh $10, $10, $13 \n\t" + "pmfhl.uw $11 \n\t" + "pinteh $10, $11, $10 \n\t" //r10 = level * qmul + "paddh $10, $10, $13 \n\t" "pand $10, $10, $12 \n\t" - "sq $10, -16($14) \n\t" - "bgez %2, 1b \n\t" - :: "r"(qmul), "r" (qadd), "r" (nCoeffs), "r" (block) : "$8", "$9", "$10", "$11", "$12", "$13", "$14", "memory" ); + "sq $10, -16($14) \n\t" + "bgez %2, 1b \n\t" + :: "r"(qmul), "r" (qadd), "r" (nCoeffs), "r" (block) : "$8", "$9", "$10", "$11", "$12", "$13", "$14", "memory" ); if(s->mb_intra) block[0]= level;
--- a/pthread.c Wed Dec 21 17:50:40 2005 +0000 +++ b/pthread.c Thu Dec 22 01:10:11 2005 +0000 @@ -53,24 +53,24 @@ pthread_mutex_lock(&c->current_job_lock); self_id = c->current_job++; for (;;){ - while (our_job >= c->job_count) { - if (c->current_job == thread_count + c->job_count) - pthread_cond_signal(&c->last_job_cond); + while (our_job >= c->job_count) { + if (c->current_job == thread_count + c->job_count) + pthread_cond_signal(&c->last_job_cond); - pthread_cond_wait(&c->current_job_cond, &c->current_job_lock); + pthread_cond_wait(&c->current_job_cond, &c->current_job_lock); our_job = self_id; - if (c->done) { - pthread_mutex_unlock(&c->current_job_lock); - return NULL; - } - } - pthread_mutex_unlock(&c->current_job_lock); + if (c->done) { + pthread_mutex_unlock(&c->current_job_lock); + return NULL; + } + } + pthread_mutex_unlock(&c->current_job_lock); - c->rets[our_job%c->rets_count] = c->func(avctx, c->args[our_job]); + c->rets[our_job%c->rets_count] = c->func(avctx, c->args[our_job]); - pthread_mutex_lock(&c->current_job_lock); - our_job = c->current_job++; + pthread_mutex_lock(&c->current_job_lock); + our_job = c->current_job++; } } @@ -116,10 +116,10 @@ c->func = func; if (ret) { c->rets = ret; - c->rets_count = job_count; + c->rets_count = job_count; } else { c->rets = &dummy_ret; - c->rets_count = 1; + c->rets_count = 1; } pthread_cond_broadcast(&c->current_job_cond); @@ -154,8 +154,8 @@ pthread_mutex_lock(&c->current_job_lock); for (i=0; i<thread_count; i++) { if(pthread_create(&c->workers[i], NULL, worker, avctx)) { - avctx->thread_count = i; - pthread_mutex_unlock(&c->current_job_lock); + avctx->thread_count = i; + pthread_mutex_unlock(&c->current_job_lock); avcodec_thread_free(avctx); return -1; }
--- a/qpeg.c Wed Dec 21 17:50:40 2005 +0000 +++ b/qpeg.c Thu Dec 22 01:10:11 2005 +0000 @@ -33,7 +33,7 @@ } QpegContext; static void qpeg_decode_intra(uint8_t *src, uint8_t *dst, int size, - int stride, int width, int height) + int stride, int width, int height) { int i; int code; @@ -47,64 +47,64 @@ dst = dst + height * stride; while((size > 0) && (rows_to_go > 0)) { - code = *src++; - size--; - run = copy = 0; - if(code == 0xFC) /* end-of-picture code */ - break; - if(code >= 0xF8) { /* very long run */ - c0 = *src++; - c1 = *src++; - size -= 2; - run = ((code & 0x7) << 16) + (c0 << 8) + c1 + 2; - } else if (code >= 0xF0) { /* long run */ - c0 = *src++; - size--; - run = ((code & 0xF) << 8) + c0 + 2; - } else if (code >= 0xE0) { /* short run */ - run = (code & 0x1F) + 2; - } else if (code >= 0xC0) { /* very long copy */ - c0 = *src++; - c1 = *src++; - size -= 2; - copy = ((code & 0x3F) << 16) + (c0 << 8) + c1 + 1; - } else if (code >= 0x80) { /* long copy */ - c0 = *src++; - size--; - copy = ((code & 0x7F) << 8) + c0 + 1; - } else { /* short copy */ - copy = code + 1; - } + code = *src++; + size--; + run = copy = 0; + if(code == 0xFC) /* end-of-picture code */ + break; + if(code >= 0xF8) { /* very long run */ + c0 = *src++; + c1 = *src++; + size -= 2; + run = ((code & 0x7) << 16) + (c0 << 8) + c1 + 2; + } else if (code >= 0xF0) { /* long run */ + c0 = *src++; + size--; + run = ((code & 0xF) << 8) + c0 + 2; + } else if (code >= 0xE0) { /* short run */ + run = (code & 0x1F) + 2; + } else if (code >= 0xC0) { /* very long copy */ + c0 = *src++; + c1 = *src++; + size -= 2; + copy = ((code & 0x3F) << 16) + (c0 << 8) + c1 + 1; + } else if (code >= 0x80) { /* long copy */ + c0 = *src++; + size--; + copy = ((code & 0x7F) << 8) + c0 + 1; + } else { /* short copy */ + copy = code + 1; + } - /* perform actual run or copy */ - if(run) { - int p; + /* perform actual run or copy */ + if(run) { + int p; - p = *src++; - size--; - for(i = 0; i < run; i++) { - dst[filled++] = p; - if (filled >= width) { - filled = 0; - dst -= stride; + p = *src++; + size--; + for(i = 0; i < run; i++) { + dst[filled++] = p; + if (filled >= width) { + filled = 0; + dst -= stride; rows_to_go--; if(rows_to_go <= 0) break; - } - } - } else { + } + } + } else { size -= copy; - for(i = 0; i < copy; i++) { - dst[filled++] = *src++; - if (filled >= width) { - filled = 0; - dst -= stride; + for(i = 0; i < copy; i++) { + dst[filled++] = *src++; + if (filled >= width) { + filled = 0; + dst -= stride; rows_to_go--; if(rows_to_go <= 0) break; - } - } - } + } + } + } } } @@ -115,8 +115,8 @@ /* Decodes delta frames */ static void qpeg_decode_inter(uint8_t *src, uint8_t *dst, int size, - int stride, int width, int height, - int delta, uint8_t *ctable, uint8_t *refdata) + int stride, int width, int height, + int delta, uint8_t *ctable, uint8_t *refdata) { int i, j; int code; @@ -126,7 +126,7 @@ /* copy prev frame */ for(i = 0; i < height; i++) - memcpy(refdata + (i * width), dst + (i * stride), width); + memcpy(refdata + (i * width), dst + (i * stride), width); orig_height = height; blkdata = src - 0x86; @@ -134,36 +134,36 @@ dst = dst + height * stride; while((size > 0) && (height >= 0)) { - code = *src++; - size--; + code = *src++; + size--; - if(delta) { - /* motion compensation */ - while((code & 0xF0) == 0xF0) { - if(delta == 1) { - int me_idx; - int me_w, me_h, me_x, me_y; - uint8_t *me_plane; - int corr, val; + if(delta) { + /* motion compensation */ + while((code & 0xF0) == 0xF0) { + if(delta == 1) { + int me_idx; + int me_w, me_h, me_x, me_y; + uint8_t *me_plane; + int corr, val; - /* get block size by index */ - me_idx = code & 0xF; - me_w = qpeg_table_w[me_idx]; - me_h = qpeg_table_h[me_idx]; + /* get block size by index */ + me_idx = code & 0xF; + me_w = qpeg_table_w[me_idx]; + me_h = qpeg_table_h[me_idx]; - /* extract motion vector */ - corr = *src++; - size--; + /* extract motion vector */ + corr = *src++; + size--; - val = corr >> 4; - if(val > 7) - val -= 16; - me_x = val; + val = corr >> 4; + if(val > 7) + val -= 16; + me_x = val; - val = corr & 0xF; - if(val > 7) - val -= 16; - me_y = val; + val = corr & 0xF; + if(val > 7) + val -= 16; + me_y = val; /* check motion vector */ if ((me_x + filled < 0) || (me_x + me_w + filled > width) || @@ -178,73 +178,73 @@ for(i = 0; i < me_w; i++) dst[filled + i - (j * stride)] = me_plane[i - (j * width)]; } - } - } - code = *src++; - size--; - } - } + } + } + code = *src++; + size--; + } + } - if(code == 0xE0) /* end-of-picture code */ - break; - if(code > 0xE0) { /* run code: 0xE1..0xFF */ - int p; + if(code == 0xE0) /* end-of-picture code */ + break; + if(code > 0xE0) { /* run code: 0xE1..0xFF */ + int p; - code &= 0x1F; - p = *src++; - size--; - for(i = 0; i <= code; i++) { - dst[filled++] = p; - if(filled >= width) { - filled = 0; - dst -= stride; - height--; - } - } - } else if(code >= 0xC0) { /* copy code: 0xC0..0xDF */ - code &= 0x1F; + code &= 0x1F; + p = *src++; + size--; + for(i = 0; i <= code; i++) { + dst[filled++] = p; + if(filled >= width) { + filled = 0; + dst -= stride; + height--; + } + } + } else if(code >= 0xC0) { /* copy code: 0xC0..0xDF */ + code &= 0x1F; - for(i = 0; i <= code; i++) { - dst[filled++] = *src++; - if(filled >= width) { - filled = 0; - dst -= stride; - height--; - } - } - size -= code + 1; - } else if(code >= 0x80) { /* skip code: 0x80..0xBF */ - int skip; + for(i = 0; i <= code; i++) { + dst[filled++] = *src++; + if(filled >= width) { + filled = 0; + dst -= stride; + height--; + } + } + size -= code + 1; + } else if(code >= 0x80) { /* skip code: 0x80..0xBF */ + int skip; - code &= 0x3F; - /* codes 0x80 and 0x81 are actually escape codes, - skip value minus constant is in the next byte */ - if(!code) - skip = (*src++) + 64; - else if(code == 1) - skip = (*src++) + 320; - else - skip = code; - filled += skip; - while( filled >= width) { - filled -= width; - dst -= stride; - height--; + code &= 0x3F; + /* codes 0x80 and 0x81 are actually escape codes, + skip value minus constant is in the next byte */ + if(!code) + skip = (*src++) + 64; + else if(code == 1) + skip = (*src++) + 320; + else + skip = code; + filled += skip; + while( filled >= width) { + filled -= width; + dst -= stride; + height--; if(height < 0) break; - } - } else { - /* zero code treated as one-pixel skip */ - if(code) - dst[filled++] = ctable[code & 0x7F]; - else - filled++; - if(filled >= width) { - filled = 0; - dst -= stride; - height--; - } - } + } + } else { + /* zero code treated as one-pixel skip */ + if(code) + dst[filled++] = ctable[code & 0x7F]; + else + filled++; + if(filled >= width) { + filled = 0; + dst -= stride; + height--; + } + } } } @@ -267,10 +267,10 @@ } outdata = a->pic.data[0]; if(buf[0x85] == 0x10) { - qpeg_decode_intra(buf+0x86, outdata, buf_size - 0x86, a->pic.linesize[0], avctx->width, avctx->height); + qpeg_decode_intra(buf+0x86, outdata, buf_size - 0x86, a->pic.linesize[0], avctx->width, avctx->height); } else { - delta = buf[0x85]; - qpeg_decode_inter(buf+0x86, outdata, buf_size - 0x86, a->pic.linesize[0], avctx->width, avctx->height, delta, buf + 4, a->refdata); + delta = buf[0x85]; + qpeg_decode_inter(buf+0x86, outdata, buf_size - 0x86, a->pic.linesize[0], avctx->width, avctx->height, delta, buf + 4, a->refdata); } /* make the palette available on the way out */
--- a/ra144.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ra144.c Thu Dec 22 01:10:11 2005 +0000 @@ -20,72 +20,72 @@ #include "avcodec.h" #include "ra144.h" -#define DATABLOCK1 20 /* size of 14.4 input block in bytes */ -#define DATACHUNK1 1440 /* size of 14.4 input chunk in bytes */ -#define AUDIOBLOCK 160 /* size of output block in 16-bit words (320 bytes) */ -#define AUDIOBUFFER 12288 /* size of output buffer in 16-bit words (24576 bytes) */ +#define DATABLOCK1 20 /* size of 14.4 input block in bytes */ +#define DATACHUNK1 1440 /* size of 14.4 input chunk in bytes */ +#define AUDIOBLOCK 160 /* size of output block in 16-bit words (320 bytes) */ +#define AUDIOBUFFER 12288 /* size of output buffer in 16-bit words (24576 bytes) */ /* consts */ -#define NBLOCKS 4 /* number of segments within a block */ -#define BLOCKSIZE 40 /* (quarter) block size in 16-bit words (80 bytes) */ -#define HALFBLOCK 20 /* BLOCKSIZE/2 */ -#define BUFFERSIZE 146 /* for do_output */ +#define NBLOCKS 4 /* number of segments within a block */ +#define BLOCKSIZE 40 /* (quarter) block size in 16-bit words (80 bytes) */ +#define HALFBLOCK 20 /* BLOCKSIZE/2 */ +#define BUFFERSIZE 146 /* for do_output */ /* internal globals */ typedef struct { - unsigned int resetflag, val, oldval; - unsigned int unpacked[28]; /* buffer for unpacked input */ - unsigned int *iptr; /* pointer to current input (from unpacked) */ - unsigned int gval; - unsigned short *gsp; - unsigned int gbuf1[8]; - unsigned short gbuf2[120]; - signed short output_buffer[40]; - unsigned int *decptr; /* decoder ptr */ - signed short *decsp; + unsigned int resetflag, val, oldval; + unsigned int unpacked[28]; /* buffer for unpacked input */ + unsigned int *iptr; /* pointer to current input (from unpacked) */ + unsigned int gval; + unsigned short *gsp; + unsigned int gbuf1[8]; + unsigned short gbuf2[120]; + signed short output_buffer[40]; + unsigned int *decptr; /* decoder ptr */ + signed short *decsp; - /* the swapped buffers */ - unsigned int swapb1a[10]; - unsigned int swapb2a[10]; - unsigned int swapb1b[10]; - unsigned int swapb2b[10]; - unsigned int *swapbuf1; - unsigned int *swapbuf2; - unsigned int *swapbuf1alt; - unsigned int *swapbuf2alt; + /* the swapped buffers */ + unsigned int swapb1a[10]; + unsigned int swapb2a[10]; + unsigned int swapb1b[10]; + unsigned int swapb2b[10]; + unsigned int *swapbuf1; + unsigned int *swapbuf2; + unsigned int *swapbuf1alt; + unsigned int *swapbuf2alt; - unsigned int buffer[5]; - unsigned short int buffer_2[148]; - unsigned short int buffer_a[40]; - unsigned short int buffer_b[40]; - unsigned short int buffer_c[40]; - unsigned short int buffer_d[40]; + unsigned int buffer[5]; + unsigned short int buffer_2[148]; + unsigned short int buffer_a[40]; + unsigned short int buffer_b[40]; + unsigned short int buffer_c[40]; + unsigned short int buffer_d[40]; - unsigned short int work[50]; - unsigned short *sptr; + unsigned short int work[50]; + unsigned short *sptr; - int buffer1[10]; - int buffer2[10]; + int buffer1[10]; + int buffer2[10]; - signed short wavtable1[2304]; - unsigned short wavtable2[2304]; + signed short wavtable1[2304]; + unsigned short wavtable2[2304]; } Real144_internal; static int ra144_decode_init(AVCodecContext * avctx) { - Real144_internal *glob=avctx->priv_data; + Real144_internal *glob=avctx->priv_data; - memset(glob,0,sizeof(Real144_internal)); - glob->resetflag=1; - glob->swapbuf1=glob->swapb1a; - glob->swapbuf2=glob->swapb2a; - glob->swapbuf1alt=glob->swapb1b; - glob->swapbuf2alt=glob->swapb2b; + memset(glob,0,sizeof(Real144_internal)); + glob->resetflag=1; + glob->swapbuf1=glob->swapb1a; + glob->swapbuf2=glob->swapb2a; + glob->swapbuf1alt=glob->swapb1b; + glob->swapbuf2alt=glob->swapb2b; - memcpy(glob->wavtable1,wavtable1,sizeof(wavtable1)); - memcpy(glob->wavtable2,wavtable2,sizeof(wavtable2)); + memcpy(glob->wavtable1,wavtable1,sizeof(wavtable1)); + memcpy(glob->wavtable2,wavtable2,sizeof(wavtable2)); - return 0; + return 0; } static void final(Real144_internal *glob, short *i1, short *i2, void *out, int *statbuf, int len);
--- a/ra288.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ra288.c Thu Dec 22 01:10:11 2005 +0000 @@ -21,23 +21,23 @@ #include "ra288.h" typedef struct { - float history[8]; - float output[40]; - float pr1[36]; - float pr2[10]; - int phase, phasep; + float history[8]; + float output[40]; + float pr1[36]; + float pr2[10]; + int phase, phasep; - float st1a[111],st1b[37],st1[37]; - float st2a[38],st2b[11],st2[11]; - float sb[41]; - float lhist[10]; + float st1a[111],st1b[37],st1[37]; + float st2a[38],st2b[11],st2[11]; + float sb[41]; + float lhist[10]; } Real288_internal; static int ra288_decode_init(AVCodecContext * avctx) { - Real288_internal *glob=avctx->priv_data; - memset(glob,0,sizeof(Real288_internal)); - return 0; + Real288_internal *glob=avctx->priv_data; + memset(glob,0,sizeof(Real288_internal)); + return 0; } static void prodsum(float *tgt, float *src, int len, int n); @@ -109,7 +109,7 @@ for (sum=32,x=10;x--;sum-=glob->pr2[x]*glob->lhist[x]); if (sum<0) sum=0; else if (sum>60) sum=60; - sumsum=exp(sum*0.1151292546497)*f; /* pow(10.0,sum/20)*f */ + sumsum=exp(sum*0.1151292546497)*f; /* pow(10.0,sum/20)*f */ for (sum=0,x=5;x--;) { buffer[x]=table[x]*sumsum; sum+=buffer[x]*buffer[x]; } if ((sum/=5)<1) sum=1; @@ -232,8 +232,8 @@ if (buf_size < avctx->block_align) { - av_log(avctx, AV_LOG_ERROR, "ffra288: Error! Input buffer is too small [%d<%d]\n",buf_size,avctx->block_align); - return 0; + av_log(avctx, AV_LOG_ERROR, "ffra288: Error! Input buffer is too small [%d<%d]\n",buf_size,avctx->block_align); + return 0; } datao = data;
--- a/ra288.h Wed Dec 21 17:50:40 2005 +0000 +++ b/ra288.h Thu Dec 22 01:10:11 2005 +0000 @@ -24,180 +24,180 @@ -0.515625,-0.90234375,-1.57910156,-2.76342773 }; static const float codetable[640]={ - 0.326171875, -1.4404296875, -0.6123046875, -0.8740234375, -1.24658203125, - -2.45703125, -2.23486328125, -0.51025390625, 1.419921875, 1.6201171875, - -1.37646484375, -1.30712890625, -0.462890625, -1.37939453125, -2.1728515625, - -3.26123046875, -0.166015625, 0.7236328125, -0.623046875, 0.6162109375, - -0.2744140625, -3.29931640625, 0.62548828125, 0.08740234375, -0.6220703125, - -1.2265625, -3.4814453125, -2.40478515625, 3.37548828125, 1.17724609375, - -1.2099609375, -0.076171875, 2.28662109375, -1.89111328125, 0, - -4.0078125, 1.044921875, -0.2333984375, -1.35986328125, 0.26025390625, - 0.92236328125, 1.34716796875, 0.67431640625, -3.39599609375, -2.88720703125, - 2.4814453125, -1.201171875, -2.8212890625, 0.87744140625, 0.27734375, - -1.078125, -1.61572265625, -2.20849609375, -3.044921875, -3.66455078125, - -1.32763671875, 2.1279296875, -1.458984375, -0.56103515625, 1.30078125, - 0.61474609375, 0.48583984375, 1.32373046875, -1.203125, -5.0732421875, - 0.8408203125, -3.69580078125, -1.3388671875, 1.06005859375, -1.13720703125, - 0.50390625, 0.36474609375, -0.4189453125, -3.8798828125, -6.27099609375, - 1.5166015625, 2.37109375, -2.04736328125, -1.24072265625, 0.50537109375, - 0.9091796875, -0.46875, -3.236328125, 0.2001953125, 2.8720703125, - -1.21728515625, -1.283203125, -1.953125, -0.029296875, 3.5166015625, - -1.3046875, 0.7060546875, 0.75, -1.87060546875, 0.60205078125, - -2.5888671875, 3.375, 0.77587890625, -2.04443359375, 1.78955078125, - -1.6875, -3.9892578125, -3.76416015625, 0.67578125, 2.2939453125, - -2.29443359375, -3.03173828125, -5.45703125, 3.95703125, 8.2177734375, - 0.4541015625, 3.419921875, 0.61962890625, -4.38330078125, 1.25341796875, - 2.27001953125, 5.763671875, 1.68017578125, -2.76220703125, 0.58544921875, - 1.2412109375, -0.08935546875, -4.32568359375, -3.89453125, 1.5771484375, - -1.40234375, -0.98193359375, -4.74267578125, -4.09423828125, 6.33935546875, - 1.5068359375, 1.044921875, -1.796875, -4.70849609375, -1.4140625, - -3.71533203125, 3.18115234375, -1.11474609375, -1.2314453125, 3.091796875, - -1.62744140625, -2.744140625, -4.4580078125, -5.43505859375, 2.70654296875, - -0.19873046875, -3.28173828125, -8.5283203125, -1.41064453125, 5.6484375, - 1.802734375, 3.318359375, -0.1279296875, -5.2958984375, -0.90625, - 3.55224609375, 6.544921875, -1.45947265625, -5.17333984375, 2.41015625, - 0.119140625, -1.08349609375, 1.296875, 1.84375, -2.642578125, - -1.97412109375, -2.8974609375, 1.04052734375, 0.42138671875, -1.3994140625, - -1.6123046875, 0.85107421875, -0.9794921875, -0.0625, -1.001953125, - -3.10595703125, 1.6318359375, -0.77294921875, -0.01025390625, 0.5576171875, - -1.87353515625, -0.89404296875, 3.12353515625, 1.24267578125, -1.390625, - -4.556640625, -3.1875, 2.59228515625, 0.9697265625, -1.09619140625, - -2.1923828125, 0.365234375, 0.94482421875, -1.47802734375, -0.24072265625, - -4.51904296875, 2.6201171875, 1.55908203125, -2.19384765625, 0.87109375, - 2.3359375, -0.1806640625, 0.9111328125, 0.51611328125, -0.92236328125, - 3.5849609375, -1.3134765625, -1.25830078125, 0.330078125, -0.29833984375, - -0.2451171875, 1.09130859375, -0.9033203125, -0.86767578125, -1.00048828125, - 0.49365234375, 1.89453125, -1.20361328125, 1.07861328125, -0.07421875, - 1.265625, 1.38134765625, 2.728515625, 1.38623046875, -3.5673828125, - -1.48876953125, -2.4013671875, 2.90771484375, 4.49267578125, -2.17138671875, - 0.34033203125, 1.908203125, 2.8310546875, -2.17333984375, -2.267578125, - -1.03564453125, 2.658203125, -1.2548828125, 0.15673828125, -0.5869140625, - 1.3896484375, -1.0185546875, 1.724609375, 0.2763671875, -0.345703125, - -2.08935546875, 0.4638671875, 2.431640625, 1.83056640625, 0.220703125, - -1.212890625, 1.7099609375, 0.83935546875, -0.0830078125, 0.1162109375, - -1.67724609375, 0.12841796875, 1.0322265625, -0.97900390625, 1.15283203125, - -3.5830078125, -0.58984375, 4.56396484375, -0.59375, -1.95947265625, - -6.5908203125, -0.21435546875, 3.919921875, -2.06640625, 0.17626953125, - -1.82080078125, 2.65283203125, 0.978515625, -2.30810546875, -0.61474609375, - -1.9462890625, 3.78076171875, 4.11572265625, -1.80224609375, -0.48193359375, - 2.5380859375, -0.20654296875, 0.5615234375, -0.62548828125, 0.3984375, - 3.61767578125, 2.00634765625, -1.92822265625, 1.3134765625, 0.0146484384313, - 0.6083984375, 1.49169921875, -0.01708984375, -0.6689453125, -0.1201171875, - -0.72705078125, 2.75146484375, -0.3310546875, -1.28271484375, 1.5478515625, - 2.3583984375, -2.23876953125, 0.98046875, -0.5185546875, 0.39013671875, - -0.06298828125, 0.35009765625, 2.2431640625, 7.29345703125, 5.2275390625, - 0.20361328125, 1.34716796875, 0.9033203125, -2.46923828125, -0.56298828125, - -1.89794921875, 3.59423828125, -2.81640625, 2.09228515625, 0.3251953125, - 0.70458984375, -0.4580078125, 0.009765625, -1.03466796875, -0.82861328125, - -1.8125, -1.6611328125, -1.080078125, 0.0537109375, 1.04296875, - -1.44140625, 0.005859375, -0.765625, -1.708984375, -0.90576171875, - -0.64208984375, -0.84521484375, 0.56640625, -0.2724609375, 0.83447265625, - 0.04296875, -2.23095703125, 0.0947265625, -0.2216796875, -1.44384765625, - -1.38623046875, -0.8134765625, -0.13330078125, 1.017578125, -0.07568359375, - -0.09228515625, -1.16015625, 0.81201171875, -0.5078125, -1.19580078125, - -1.3876953125, -0.66845703125, 0.310546875, -0.12109375, -1.30712890625, - 0.74072265625, 0.03857421875, -1.47119140625, -1.79150390625, -0.47509765625, - 0.93408203125, -1.21728515625, -2.59375, -0.36572265625, 0.62060546875, - -1.41748046875, -1.623046875, -1.833984375, -1.8017578125, -0.89306640625, - -1.42236328125, -0.75537109375, -1.34765625, -0.6865234375, 0.548828125, - 0.900390625, -0.8955078125, 0.22265625, 0.3447265625, -2.0859375, - 0.22802734375, -2.078125, -0.93212890625, 0.74267578125, 0.5537109375, - -0.06201171875, -0.4853515625, -0.31103515625, -0.72802734375, -3.1708984375, - 0.42626953125, -0.99853515625, -1.869140625, -1.36328125, -0.2822265625, - 1.12841796875, -0.88720703125, 1.28515625, -1.490234375, 0.9609375, - 0.31298828125, 0.5830078125, 0.92431640625, 2.00537109375, 3.0966796875, - -0.02197265625, 0.5849609375, 1.0546875, -0.70751953125, 1.07568359375, - -0.978515625, 0.83642578125, 1.7177734375, 1.294921875, 2.07568359375, - 1.43359375, -1.9375, 0.625, 0.06396484375, -0.720703125, - 1.38037109375, 0.00390625, -0.94140625, 1.2978515625, 1.71533203125, - 1.56201171875, -0.3984375, 1.31201171875, -0.85009765625, -0.68701171875, - 1.439453125, 1.96728515625, 0.1923828125, -0.12353515625, 0.6337890625, - 2.0927734375, 0.02490234375, -2.20068359375, -0.015625, -0.32177734375, - 1.90576171875, 2.7568359375, -2.728515625, -1.265625, 2.78662109375, - -0.2958984375, 0.6025390625, -0.78466796875, -2.53271484375, 0.32421875, - -0.25634765625, 1.767578125, -1.0703125, -1.23388671875, 0.83349609375, - 2.09814453125, -1.58740234375, -1.11474609375, 0.396484375, -1.10546875, - 2.81494140625, 0.2578125, -1.60498046875, 0.66015625, 0.81640625, - 1.33544921875, 0.60595703125, -0.53857421875, -1.59814453125, -1.66357421875, - 1.96923828125, 0.8046875, -1.44775390625, -0.5732421875, 0.705078125, - 0.0361328125, 0.4482421875, 0.97607421875, 0.44677734375, -0.5009765625, - -1.21875, -0.78369140625, 0.9931640625, 1.4404296875, 0.11181640625, - -1.05859375, 0.99462890625, 0.00732421921566,-0.6171875, -0.1015625, - -1.734375, 0.7470703125, 0.28369140625, 0.72802734375, 0.4697265625, - -1.27587890625, -1.1416015625, 1.76806640625, -0.7265625, -1.06689453125, - -0.85302734375, 0.03955078125, 2.7041015625, 0.69921875, -1.10205078125, - -0.49755859375, 0.42333984375, 0.1044921875, -1.115234375, -0.7373046875, - -0.822265625, 1.375, -0.11181640625, 1.24560546875, -0.67822265625, - 1.32177734375, 0.24609375, 0.23388671875, 1.35888671875, -0.49267578125, - 1.22900390625, -0.72607421875, -0.779296875, 0.30322265625, 0.94189453125, - -0.072265625, 1.0771484375, -2.09375, 0.630859375, -0.68408203125, - -0.25732421875, 0.60693359375, -1.33349609375, 0.93212890625, 0.625, - 1.04931640625, -0.73291015625, 1.80078125, 0.2978515625, -2.24169921875, - 1.6142578125, -1.64501953125, 0.91552734375, 1.775390625, -0.59423828125, - 1.2568359375, 1.22705078125, 0.70751953125, -1.5009765625, -2.43115234375, - 0.3974609375, 0.8916015625, -1.21923828125, 2.0673828125, -1.99072265625, - 0.8125, -0.107421875, 1.6689453125, 0.4892578125, 0.54443359375, - 0.38134765625, 0.8095703125, 1.91357421875, 2.9931640625, 1.533203125, - 0.560546875, 1.98486328125, 0.740234375, 0.39794921875, 0.09716796875, - 0.58154296875, 1.21533203125, 1.25048828125, 1.18212890625, 1.19287109375, - 0.3759765625, -2.88818359375, 2.69287109375, -0.1796875, -1.56201171875, - 0.5810546875, 0.51123046875, 1.8271484375, 3.38232421875, -1.02001953125, - 0.142578125, 1.51318359375, 2.103515625, -0.3701171875, -1.19873046875, - 0.25537109375, 1.91455078125, 1.974609375, 0.6767578125, 0.04150390625, - 2.13232421875, 0.4912109375, -0.611328125, -0.7158203125, -0.67529296875, - 1.880859375, 0.77099609375, -0.03759765625, 1.0078125, 0.423828125, - 2.49462890625, 1.42529296875, -0.0986328125, 0.17529296875, -0.24853515625, - 1.7822265625, 1.5654296875, 1.12451171875, 0.82666015625, 0.6328125, - 1.41845703125, -1.90771484375, 0.11181640625, -0.583984375, -1.138671875, - 2.91845703125, -1.75048828125, 0.39306640625, 1.86767578125, -1.5322265625, - 1.8291015625, -0.2958984375, 0.02587890625, -0.13134765625, -1.61181640625, - 0.2958984375, 0.9853515625, -0.642578125, 1.984375, 0.1943359375 + 0.326171875, -1.4404296875, -0.6123046875, -0.8740234375, -1.24658203125, + -2.45703125, -2.23486328125, -0.51025390625, 1.419921875, 1.6201171875, + -1.37646484375, -1.30712890625, -0.462890625, -1.37939453125, -2.1728515625, + -3.26123046875, -0.166015625, 0.7236328125, -0.623046875, 0.6162109375, + -0.2744140625, -3.29931640625, 0.62548828125, 0.08740234375, -0.6220703125, + -1.2265625, -3.4814453125, -2.40478515625, 3.37548828125, 1.17724609375, + -1.2099609375, -0.076171875, 2.28662109375, -1.89111328125, 0, + -4.0078125, 1.044921875, -0.2333984375, -1.35986328125, 0.26025390625, + 0.92236328125, 1.34716796875, 0.67431640625, -3.39599609375, -2.88720703125, + 2.4814453125, -1.201171875, -2.8212890625, 0.87744140625, 0.27734375, + -1.078125, -1.61572265625, -2.20849609375, -3.044921875, -3.66455078125, + -1.32763671875, 2.1279296875, -1.458984375, -0.56103515625, 1.30078125, + 0.61474609375, 0.48583984375, 1.32373046875, -1.203125, -5.0732421875, + 0.8408203125, -3.69580078125, -1.3388671875, 1.06005859375, -1.13720703125, + 0.50390625, 0.36474609375, -0.4189453125, -3.8798828125, -6.27099609375, + 1.5166015625, 2.37109375, -2.04736328125, -1.24072265625, 0.50537109375, + 0.9091796875, -0.46875, -3.236328125, 0.2001953125, 2.8720703125, + -1.21728515625, -1.283203125, -1.953125, -0.029296875, 3.5166015625, + -1.3046875, 0.7060546875, 0.75, -1.87060546875, 0.60205078125, + -2.5888671875, 3.375, 0.77587890625, -2.04443359375, 1.78955078125, + -1.6875, -3.9892578125, -3.76416015625, 0.67578125, 2.2939453125, + -2.29443359375, -3.03173828125, -5.45703125, 3.95703125, 8.2177734375, + 0.4541015625, 3.419921875, 0.61962890625, -4.38330078125, 1.25341796875, + 2.27001953125, 5.763671875, 1.68017578125, -2.76220703125, 0.58544921875, + 1.2412109375, -0.08935546875, -4.32568359375, -3.89453125, 1.5771484375, + -1.40234375, -0.98193359375, -4.74267578125, -4.09423828125, 6.33935546875, + 1.5068359375, 1.044921875, -1.796875, -4.70849609375, -1.4140625, + -3.71533203125, 3.18115234375, -1.11474609375, -1.2314453125, 3.091796875, + -1.62744140625, -2.744140625, -4.4580078125, -5.43505859375, 2.70654296875, + -0.19873046875, -3.28173828125, -8.5283203125, -1.41064453125, 5.6484375, + 1.802734375, 3.318359375, -0.1279296875, -5.2958984375, -0.90625, + 3.55224609375, 6.544921875, -1.45947265625, -5.17333984375, 2.41015625, + 0.119140625, -1.08349609375, 1.296875, 1.84375, -2.642578125, + -1.97412109375, -2.8974609375, 1.04052734375, 0.42138671875, -1.3994140625, + -1.6123046875, 0.85107421875, -0.9794921875, -0.0625, -1.001953125, + -3.10595703125, 1.6318359375, -0.77294921875, -0.01025390625, 0.5576171875, + -1.87353515625, -0.89404296875, 3.12353515625, 1.24267578125, -1.390625, + -4.556640625, -3.1875, 2.59228515625, 0.9697265625, -1.09619140625, + -2.1923828125, 0.365234375, 0.94482421875, -1.47802734375, -0.24072265625, + -4.51904296875, 2.6201171875, 1.55908203125, -2.19384765625, 0.87109375, + 2.3359375, -0.1806640625, 0.9111328125, 0.51611328125, -0.92236328125, + 3.5849609375, -1.3134765625, -1.25830078125, 0.330078125, -0.29833984375, + -0.2451171875, 1.09130859375, -0.9033203125, -0.86767578125, -1.00048828125, + 0.49365234375, 1.89453125, -1.20361328125, 1.07861328125, -0.07421875, + 1.265625, 1.38134765625, 2.728515625, 1.38623046875, -3.5673828125, + -1.48876953125, -2.4013671875, 2.90771484375, 4.49267578125, -2.17138671875, + 0.34033203125, 1.908203125, 2.8310546875, -2.17333984375, -2.267578125, + -1.03564453125, 2.658203125, -1.2548828125, 0.15673828125, -0.5869140625, + 1.3896484375, -1.0185546875, 1.724609375, 0.2763671875, -0.345703125, + -2.08935546875, 0.4638671875, 2.431640625, 1.83056640625, 0.220703125, + -1.212890625, 1.7099609375, 0.83935546875, -0.0830078125, 0.1162109375, + -1.67724609375, 0.12841796875, 1.0322265625, -0.97900390625, 1.15283203125, + -3.5830078125, -0.58984375, 4.56396484375, -0.59375, -1.95947265625, + -6.5908203125, -0.21435546875, 3.919921875, -2.06640625, 0.17626953125, + -1.82080078125, 2.65283203125, 0.978515625, -2.30810546875, -0.61474609375, + -1.9462890625, 3.78076171875, 4.11572265625, -1.80224609375, -0.48193359375, + 2.5380859375, -0.20654296875, 0.5615234375, -0.62548828125, 0.3984375, + 3.61767578125, 2.00634765625, -1.92822265625, 1.3134765625, 0.0146484384313, + 0.6083984375, 1.49169921875, -0.01708984375, -0.6689453125, -0.1201171875, + -0.72705078125, 2.75146484375, -0.3310546875, -1.28271484375, 1.5478515625, + 2.3583984375, -2.23876953125, 0.98046875, -0.5185546875, 0.39013671875, + -0.06298828125, 0.35009765625, 2.2431640625, 7.29345703125, 5.2275390625, + 0.20361328125, 1.34716796875, 0.9033203125, -2.46923828125, -0.56298828125, + -1.89794921875, 3.59423828125, -2.81640625, 2.09228515625, 0.3251953125, + 0.70458984375, -0.4580078125, 0.009765625, -1.03466796875, -0.82861328125, + -1.8125, -1.6611328125, -1.080078125, 0.0537109375, 1.04296875, + -1.44140625, 0.005859375, -0.765625, -1.708984375, -0.90576171875, + -0.64208984375, -0.84521484375, 0.56640625, -0.2724609375, 0.83447265625, + 0.04296875, -2.23095703125, 0.0947265625, -0.2216796875, -1.44384765625, + -1.38623046875, -0.8134765625, -0.13330078125, 1.017578125, -0.07568359375, + -0.09228515625, -1.16015625, 0.81201171875, -0.5078125, -1.19580078125, + -1.3876953125, -0.66845703125, 0.310546875, -0.12109375, -1.30712890625, + 0.74072265625, 0.03857421875, -1.47119140625, -1.79150390625, -0.47509765625, + 0.93408203125, -1.21728515625, -2.59375, -0.36572265625, 0.62060546875, + -1.41748046875, -1.623046875, -1.833984375, -1.8017578125, -0.89306640625, + -1.42236328125, -0.75537109375, -1.34765625, -0.6865234375, 0.548828125, + 0.900390625, -0.8955078125, 0.22265625, 0.3447265625, -2.0859375, + 0.22802734375, -2.078125, -0.93212890625, 0.74267578125, 0.5537109375, + -0.06201171875, -0.4853515625, -0.31103515625, -0.72802734375, -3.1708984375, + 0.42626953125, -0.99853515625, -1.869140625, -1.36328125, -0.2822265625, + 1.12841796875, -0.88720703125, 1.28515625, -1.490234375, 0.9609375, + 0.31298828125, 0.5830078125, 0.92431640625, 2.00537109375, 3.0966796875, + -0.02197265625, 0.5849609375, 1.0546875, -0.70751953125, 1.07568359375, + -0.978515625, 0.83642578125, 1.7177734375, 1.294921875, 2.07568359375, + 1.43359375, -1.9375, 0.625, 0.06396484375, -0.720703125, + 1.38037109375, 0.00390625, -0.94140625, 1.2978515625, 1.71533203125, + 1.56201171875, -0.3984375, 1.31201171875, -0.85009765625, -0.68701171875, + 1.439453125, 1.96728515625, 0.1923828125, -0.12353515625, 0.6337890625, + 2.0927734375, 0.02490234375, -2.20068359375, -0.015625, -0.32177734375, + 1.90576171875, 2.7568359375, -2.728515625, -1.265625, 2.78662109375, + -0.2958984375, 0.6025390625, -0.78466796875, -2.53271484375, 0.32421875, + -0.25634765625, 1.767578125, -1.0703125, -1.23388671875, 0.83349609375, + 2.09814453125, -1.58740234375, -1.11474609375, 0.396484375, -1.10546875, + 2.81494140625, 0.2578125, -1.60498046875, 0.66015625, 0.81640625, + 1.33544921875, 0.60595703125, -0.53857421875, -1.59814453125, -1.66357421875, + 1.96923828125, 0.8046875, -1.44775390625, -0.5732421875, 0.705078125, + 0.0361328125, 0.4482421875, 0.97607421875, 0.44677734375, -0.5009765625, + -1.21875, -0.78369140625, 0.9931640625, 1.4404296875, 0.11181640625, + -1.05859375, 0.99462890625, 0.00732421921566,-0.6171875, -0.1015625, + -1.734375, 0.7470703125, 0.28369140625, 0.72802734375, 0.4697265625, + -1.27587890625, -1.1416015625, 1.76806640625, -0.7265625, -1.06689453125, + -0.85302734375, 0.03955078125, 2.7041015625, 0.69921875, -1.10205078125, + -0.49755859375, 0.42333984375, 0.1044921875, -1.115234375, -0.7373046875, + -0.822265625, 1.375, -0.11181640625, 1.24560546875, -0.67822265625, + 1.32177734375, 0.24609375, 0.23388671875, 1.35888671875, -0.49267578125, + 1.22900390625, -0.72607421875, -0.779296875, 0.30322265625, 0.94189453125, + -0.072265625, 1.0771484375, -2.09375, 0.630859375, -0.68408203125, + -0.25732421875, 0.60693359375, -1.33349609375, 0.93212890625, 0.625, + 1.04931640625, -0.73291015625, 1.80078125, 0.2978515625, -2.24169921875, + 1.6142578125, -1.64501953125, 0.91552734375, 1.775390625, -0.59423828125, + 1.2568359375, 1.22705078125, 0.70751953125, -1.5009765625, -2.43115234375, + 0.3974609375, 0.8916015625, -1.21923828125, 2.0673828125, -1.99072265625, + 0.8125, -0.107421875, 1.6689453125, 0.4892578125, 0.54443359375, + 0.38134765625, 0.8095703125, 1.91357421875, 2.9931640625, 1.533203125, + 0.560546875, 1.98486328125, 0.740234375, 0.39794921875, 0.09716796875, + 0.58154296875, 1.21533203125, 1.25048828125, 1.18212890625, 1.19287109375, + 0.3759765625, -2.88818359375, 2.69287109375, -0.1796875, -1.56201171875, + 0.5810546875, 0.51123046875, 1.8271484375, 3.38232421875, -1.02001953125, + 0.142578125, 1.51318359375, 2.103515625, -0.3701171875, -1.19873046875, + 0.25537109375, 1.91455078125, 1.974609375, 0.6767578125, 0.04150390625, + 2.13232421875, 0.4912109375, -0.611328125, -0.7158203125, -0.67529296875, + 1.880859375, 0.77099609375, -0.03759765625, 1.0078125, 0.423828125, + 2.49462890625, 1.42529296875, -0.0986328125, 0.17529296875, -0.24853515625, + 1.7822265625, 1.5654296875, 1.12451171875, 0.82666015625, 0.6328125, + 1.41845703125, -1.90771484375, 0.11181640625, -0.583984375, -1.138671875, + 2.91845703125, -1.75048828125, 0.39306640625, 1.86767578125, -1.5322265625, + 1.8291015625, -0.2958984375, 0.02587890625, -0.13134765625, -1.61181640625, + 0.2958984375, 0.9853515625, -0.642578125, 1.984375, 0.1943359375 }; static const float table1[111]={ - 0.576690972, 0.580838025, 0.585013986, 0.589219987, 0.59345597, 0.597723007, - 0.602020264, 0.606384277, 0.610748291, 0.615142822, 0.619598389, 0.624084473, - 0.628570557, 0.633117676, 0.637695313, 0.642272949, 0.646911621, 0.651580811, - 0.656280518, 0.66104126, 0.665802002, 0.670593262, 0.675445557, 0.680328369, - 0.685241699, 0.690185547, 0.695159912, 0.700164795, 0.705230713, 0.710327148, - 0.715454102, 0.720611572, 0.725830078, 0.731048584, 0.736328125, 0.741638184, - 0.747009277, 0.752380371, 0.7578125, 0.763305664, 0.768798828, 0.774353027, - 0.779937744, 0.785583496, 0.791229248, 0.796936035, 0.802703857, 0.808502197, - 0.814331055, 0.820220947, 0.826141357, 0.832092285, 0.838104248, 0.844146729, - 0.850250244, 0.856384277, 0.862548828, 0.868774414, 0.875061035, 0.881378174, - 0.88772583, 0.894134521, 0.900604248, 0.907104492, 0.913635254, 0.920227051, - 0.926879883, 0.933563232, 0.940307617, 0.94708252, 0.953918457, 0.96081543, - 0.96774292, 0.974731445, 0.981781006, 0.988861084, 0.994842529, 0.998565674, - 0.999969482, 0.99911499, 0.996002197, 0.990600586, 0.982910156, 0.973022461, - 0.960876465, 0.946533203, 0.930053711, 0.911437988, 0.89074707, 0.868041992, - 0.843322754, 0.816680908, 0.788208008, 0.757904053, 0.725891113, 0.692199707, - 0.656921387, 0.620178223, 0.582000732, 0.542480469, 0.501739502, 0.459838867, - 0.416900635, 0.373016357, 0.328277588, 0.282775879, 0.236663818, 0.189971924, - 0.142852783, 0.0954284668, 0.0477600098 + 0.576690972, 0.580838025, 0.585013986, 0.589219987, 0.59345597, 0.597723007, + 0.602020264, 0.606384277, 0.610748291, 0.615142822, 0.619598389, 0.624084473, + 0.628570557, 0.633117676, 0.637695313, 0.642272949, 0.646911621, 0.651580811, + 0.656280518, 0.66104126, 0.665802002, 0.670593262, 0.675445557, 0.680328369, + 0.685241699, 0.690185547, 0.695159912, 0.700164795, 0.705230713, 0.710327148, + 0.715454102, 0.720611572, 0.725830078, 0.731048584, 0.736328125, 0.741638184, + 0.747009277, 0.752380371, 0.7578125, 0.763305664, 0.768798828, 0.774353027, + 0.779937744, 0.785583496, 0.791229248, 0.796936035, 0.802703857, 0.808502197, + 0.814331055, 0.820220947, 0.826141357, 0.832092285, 0.838104248, 0.844146729, + 0.850250244, 0.856384277, 0.862548828, 0.868774414, 0.875061035, 0.881378174, + 0.88772583, 0.894134521, 0.900604248, 0.907104492, 0.913635254, 0.920227051, + 0.926879883, 0.933563232, 0.940307617, 0.94708252, 0.953918457, 0.96081543, + 0.96774292, 0.974731445, 0.981781006, 0.988861084, 0.994842529, 0.998565674, + 0.999969482, 0.99911499, 0.996002197, 0.990600586, 0.982910156, 0.973022461, + 0.960876465, 0.946533203, 0.930053711, 0.911437988, 0.89074707, 0.868041992, + 0.843322754, 0.816680908, 0.788208008, 0.757904053, 0.725891113, 0.692199707, + 0.656921387, 0.620178223, 0.582000732, 0.542480469, 0.501739502, 0.459838867, + 0.416900635, 0.373016357, 0.328277588, 0.282775879, 0.236663818, 0.189971924, + 0.142852783, 0.0954284668, 0.0477600098 }; static const float table2[38]={ - 0.505699992, 0.524200022, 0.54339999, 0.563300014, 0.583953857, 0.60534668, - 0.627502441, 0.650482178, 0.674316406, 0.699005127, 0.724578857, 0.75112915, - 0.778625488, 0.807128906, 0.836669922, 0.86730957, 0.899078369, 0.932006836, - 0.961486816, 0.982757568, 0.995635986, 1, 0.995819092, 0.983154297, - 0.96206665, 0.932769775, 0.895507813, 0.850585938, 0.798400879, 0.739379883, - 0.674072266, 0.602996826, 0.526763916, 0.446014404, 0.361480713, 0.273834229, - 0.183868408, 0.0923461914 + 0.505699992, 0.524200022, 0.54339999, 0.563300014, 0.583953857, 0.60534668, + 0.627502441, 0.650482178, 0.674316406, 0.699005127, 0.724578857, 0.75112915, + 0.778625488, 0.807128906, 0.836669922, 0.86730957, 0.899078369, 0.932006836, + 0.961486816, 0.982757568, 0.995635986, 1, 0.995819092, 0.983154297, + 0.96206665, 0.932769775, 0.895507813, 0.850585938, 0.798400879, 0.739379883, + 0.674072266, 0.602996826, 0.526763916, 0.446014404, 0.361480713, 0.273834229, + 0.183868408, 0.0923461914 }; static const float table1a[36]={ - 0.98828125, 0.976699829, 0.965254128, 0.953942537, 0.942763507, 0.931715488, - 0.920796931, 0.910006344, 0.899342179, 0.888803005, 0.878387332, 0.868093729, - 0.857920766, 0.847867012, 0.837931097, 0.828111589, 0.818407178, 0.808816493, - 0.799338162, 0.789970934, 0.780713439, 0.771564424, 0.762522638, 0.753586829, - 0.744755745, 0.736028135, 0.727402806, 0.718878567, 0.710454226, 0.702128589, - 0.693900526, 0.685768902, 0.677732527, 0.669790328, 0.66194123, 0.654184103 + 0.98828125, 0.976699829, 0.965254128, 0.953942537, 0.942763507, 0.931715488, + 0.920796931, 0.910006344, 0.899342179, 0.888803005, 0.878387332, 0.868093729, + 0.857920766, 0.847867012, 0.837931097, 0.828111589, 0.818407178, 0.808816493, + 0.799338162, 0.789970934, 0.780713439, 0.771564424, 0.762522638, 0.753586829, + 0.744755745, 0.736028135, 0.727402806, 0.718878567, 0.710454226, 0.702128589, + 0.693900526, 0.685768902, 0.677732527, 0.669790328, 0.66194123, 0.654184103 }; static const float table2a[10]={ - 0.90625, 0.821289063, 0.74432373, 0.674499512, 0.61126709, - 0.553955078, 0.50201416, 0.454956055, 0.41229248, 0.373657227 + 0.90625, 0.821289063, 0.74432373, 0.674499512, 0.61126709, + 0.553955078, 0.50201416, 0.454956055, 0.41229248, 0.373657227 }; #endif /* RA288TABLES_H */
--- a/raw.c Wed Dec 21 17:50:40 2005 +0000 +++ b/raw.c Thu Dec 22 01:10:11 2005 +0000 @@ -70,8 +70,8 @@ const PixelFormatTag * tags = pixelFormatTags; while (tags->pix_fmt >= 0) { if (tags->pix_fmt == fmt) - return tags->fourcc; - tags++; + return tags->fourcc; + tags++; } return 0; } @@ -115,8 +115,8 @@ } static int raw_decode(AVCodecContext *avctx, - void *data, int *data_size, - uint8_t *buf, int buf_size) + void *data, int *data_size, + uint8_t *buf, int buf_size) { RawVideoContext *context = avctx->priv_data; int bytesNeeded; @@ -167,7 +167,7 @@ } static int raw_encode(AVCodecContext *avctx, - unsigned char *frame, int buf_size, void *data) + unsigned char *frame, int buf_size, void *data) { return avpicture_layout((AVPicture *)data, avctx->pix_fmt, avctx->width, avctx->height, frame, buf_size);
--- a/resample.c Wed Dec 21 17:50:40 2005 +0000 +++ b/resample.c Thu Dec 22 01:10:11 2005 +0000 @@ -131,15 +131,15 @@ if ( input_channels > 2) { - av_log(NULL, AV_LOG_ERROR, "Resampling with input channels greater than 2 unsupported."); - return NULL; + av_log(NULL, AV_LOG_ERROR, "Resampling with input channels greater than 2 unsupported."); + return NULL; } s = av_mallocz(sizeof(ReSampleContext)); if (!s) { - av_log(NULL, AV_LOG_ERROR, "Can't allocate memory for resample context."); - return NULL; + av_log(NULL, AV_LOG_ERROR, "Can't allocate memory for resample context."); + return NULL; } s->ratio = (float)output_rate / (float)input_rate;
--- a/rv10.c Wed Dec 21 17:50:40 2005 +0000 +++ b/rv10.c Thu Dec 22 01:10:11 2005 +0000 @@ -236,31 +236,31 @@ align_put_bits(&s->pb); - put_bits(&s->pb, 1, 1); /* marker */ + put_bits(&s->pb, 1, 1); /* marker */ put_bits(&s->pb, 1, (s->pict_type == P_TYPE)); - put_bits(&s->pb, 1, 0); /* not PB frame */ + put_bits(&s->pb, 1, 0); /* not PB frame */ put_bits(&s->pb, 5, s->qscale); if (s->pict_type == I_TYPE) { - /* specific MPEG like DC coding not used */ + /* specific MPEG like DC coding not used */ } /* if multiple packets per frame are sent, the position at which to display the macro blocks is coded here */ if(!full_frame){ - put_bits(&s->pb, 6, 0); /* mb_x */ - put_bits(&s->pb, 6, 0); /* mb_y */ + put_bits(&s->pb, 6, 0); /* mb_x */ + put_bits(&s->pb, 6, 0); /* mb_y */ put_bits(&s->pb, 12, s->mb_width * s->mb_height); } - put_bits(&s->pb, 3, 0); /* ignored */ + put_bits(&s->pb, 3, 0); /* ignored */ } void rv20_encode_picture_header(MpegEncContext *s, int picture_number){ put_bits(&s->pb, 2, s->pict_type); //I 0 vs. 1 ? - put_bits(&s->pb, 1, 0); /* unknown bit */ + put_bits(&s->pb, 1, 0); /* unknown bit */ put_bits(&s->pb, 5, s->qscale); put_bits(&s->pb, 8, picture_number&0xFF); //FIXME wrong, but correct is not known @@ -354,15 +354,15 @@ mb_xy= s->mb_x + s->mb_y*s->mb_width; if(show_bits(&s->gb, 12)==0 || (mb_xy && mb_xy < s->mb_num)){ - s->mb_x = get_bits(&s->gb, 6); /* mb_x */ - s->mb_y = get_bits(&s->gb, 6); /* mb_y */ + s->mb_x = get_bits(&s->gb, 6); /* mb_x */ + s->mb_y = get_bits(&s->gb, 6); /* mb_y */ mb_count = get_bits(&s->gb, 12); } else { s->mb_x = 0; s->mb_y = 0; mb_count = s->mb_width * s->mb_height; } - unk= get_bits(&s->gb, 3); /* ignored */ + unk= get_bits(&s->gb, 3); /* ignored */ //printf("%d\n", unk); s->f_code = 1; s->unrestricted_mv = 1;
--- a/sh4/dsputil_align.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sh4/dsputil_align.c Thu Dec 22 01:10:11 2005 +0000 @@ -23,249 +23,249 @@ #include "../dsputil.h" -#define LP(p) *(uint32_t*)(p) +#define LP(p) *(uint32_t*)(p) -#define UNPACK(ph,pl,tt0,tt1) do { \ - uint32_t t0,t1; t0=tt0;t1=tt1; \ - ph = ( (t0 & ~BYTE_VEC32(0x03))>>2) + ( (t1 & ~BYTE_VEC32(0x03))>>2); \ - pl = (t0 & BYTE_VEC32(0x03)) + (t1 & BYTE_VEC32(0x03)); } while(0) +#define UNPACK(ph,pl,tt0,tt1) do { \ + uint32_t t0,t1; t0=tt0;t1=tt1; \ + ph = ( (t0 & ~BYTE_VEC32(0x03))>>2) + ( (t1 & ~BYTE_VEC32(0x03))>>2); \ + pl = (t0 & BYTE_VEC32(0x03)) + (t1 & BYTE_VEC32(0x03)); } while(0) -#define rnd_PACK(ph,pl,nph,npl) ph + nph + (((pl + npl + BYTE_VEC32(0x02))>>2) & BYTE_VEC32(0x03)) -#define no_rnd_PACK(ph,pl,nph,npl) ph + nph + (((pl + npl + BYTE_VEC32(0x01))>>2) & BYTE_VEC32(0x03)) +#define rnd_PACK(ph,pl,nph,npl) ph + nph + (((pl + npl + BYTE_VEC32(0x02))>>2) & BYTE_VEC32(0x03)) +#define no_rnd_PACK(ph,pl,nph,npl) ph + nph + (((pl + npl + BYTE_VEC32(0x01))>>2) & BYTE_VEC32(0x03)) /* little endian */ -#define MERGE1(a,b,ofs) (ofs==0)?a:( ((a)>>(8*ofs))|((b)<<(32-8*ofs)) ) -#define MERGE2(a,b,ofs) (ofs==3)?b:( ((a)>>(8*(ofs+1)))|((b)<<(32-8*(ofs+1))) ) +#define MERGE1(a,b,ofs) (ofs==0)?a:( ((a)>>(8*ofs))|((b)<<(32-8*ofs)) ) +#define MERGE2(a,b,ofs) (ofs==3)?b:( ((a)>>(8*(ofs+1)))|((b)<<(32-8*(ofs+1))) ) /* big -#define MERGE1(a,b,ofs) (ofs==0)?a:( ((a)<<(8*ofs))|((b)>>(32-8*ofs)) ) -#define MERGE2(a,b,ofs) (ofs==3)?b:( ((a)<<(8+8*ofs))|((b)>>(32-8-8*ofs)) ) +#define MERGE1(a,b,ofs) (ofs==0)?a:( ((a)<<(8*ofs))|((b)>>(32-8*ofs)) ) +#define MERGE2(a,b,ofs) (ofs==3)?b:( ((a)<<(8+8*ofs))|((b)>>(32-8-8*ofs)) ) */ -#define put(d,s) d = s -#define avg(d,s) d = rnd_avg32(s,d) +#define put(d,s) d = s +#define avg(d,s) d = rnd_avg32(s,d) -#define OP_C4(ofs) \ - ref-=ofs; \ - do { \ - OP(LP(dest),MERGE1(LP(ref),LP(ref+4),ofs)); \ - ref+=stride; \ - dest+=stride; \ - } while(--height) +#define OP_C4(ofs) \ + ref-=ofs; \ + do { \ + OP(LP(dest),MERGE1(LP(ref),LP(ref+4),ofs)); \ + ref+=stride; \ + dest+=stride; \ + } while(--height) -#define OP_C40() \ - do { \ - OP(LP(dest),LP(ref)); \ - ref+=stride; \ - dest+=stride; \ - } while(--height) +#define OP_C40() \ + do { \ + OP(LP(dest),LP(ref)); \ + ref+=stride; \ + dest+=stride; \ + } while(--height) -#define OP put +#define OP put static void put_pixels4_c(uint8_t *dest,const uint8_t *ref, const int stride,int height) { - switch((int)ref&3){ - case 0: OP_C40(); return; - case 1: OP_C4(1); return; - case 2: OP_C4(2); return; - case 3: OP_C4(3); return; - } + switch((int)ref&3){ + case 0: OP_C40(); return; + case 1: OP_C4(1); return; + case 2: OP_C4(2); return; + case 3: OP_C4(3); return; + } } -#undef OP -#define OP avg +#undef OP +#define OP avg static void avg_pixels4_c(uint8_t *dest,const uint8_t *ref, const int stride,int height) { - switch((int)ref&3){ - case 0: OP_C40(); return; - case 1: OP_C4(1); return; - case 2: OP_C4(2); return; - case 3: OP_C4(3); return; - } + switch((int)ref&3){ + case 0: OP_C40(); return; + case 1: OP_C4(1); return; + case 2: OP_C4(2); return; + case 3: OP_C4(3); return; + } } -#undef OP +#undef OP -#define OP_C(ofs,sz,avg2) \ +#define OP_C(ofs,sz,avg2) \ { \ - ref-=ofs; \ - do { \ - uint32_t t0,t1; \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ - OP(LP(dest+0), MERGE1(t0,t1,ofs)); \ - t0 = LP(ref+8); \ - OP(LP(dest+4), MERGE1(t1,t0,ofs)); \ + ref-=ofs; \ + do { \ + uint32_t t0,t1; \ + t0 = LP(ref+0); \ + t1 = LP(ref+4); \ + OP(LP(dest+0), MERGE1(t0,t1,ofs)); \ + t0 = LP(ref+8); \ + OP(LP(dest+4), MERGE1(t1,t0,ofs)); \ if (sz==16) { \ - t1 = LP(ref+12); \ - OP(LP(dest+8), MERGE1(t0,t1,ofs)); \ - t0 = LP(ref+16); \ - OP(LP(dest+12), MERGE1(t1,t0,ofs)); \ + t1 = LP(ref+12); \ + OP(LP(dest+8), MERGE1(t0,t1,ofs)); \ + t0 = LP(ref+16); \ + OP(LP(dest+12), MERGE1(t1,t0,ofs)); \ } \ - ref+=stride; \ - dest+= stride; \ - } while(--height); \ + ref+=stride; \ + dest+= stride; \ + } while(--height); \ } /* aligned */ -#define OP_C0(sz,avg2) \ +#define OP_C0(sz,avg2) \ { \ - do { \ - OP(LP(dest+0), LP(ref+0)); \ - OP(LP(dest+4), LP(ref+4)); \ + do { \ + OP(LP(dest+0), LP(ref+0)); \ + OP(LP(dest+4), LP(ref+4)); \ if (sz==16) { \ - OP(LP(dest+8), LP(ref+8)); \ - OP(LP(dest+12), LP(ref+12)); \ + OP(LP(dest+8), LP(ref+8)); \ + OP(LP(dest+12), LP(ref+12)); \ } \ - ref+=stride; \ - dest+= stride; \ - } while(--height); \ + ref+=stride; \ + dest+= stride; \ + } while(--height); \ } -#define OP_X(ofs,sz,avg2) \ +#define OP_X(ofs,sz,avg2) \ { \ - ref-=ofs; \ - do { \ - uint32_t t0,t1; \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ - OP(LP(dest+0), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ - t0 = LP(ref+8); \ - OP(LP(dest+4), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ + ref-=ofs; \ + do { \ + uint32_t t0,t1; \ + t0 = LP(ref+0); \ + t1 = LP(ref+4); \ + OP(LP(dest+0), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ + t0 = LP(ref+8); \ + OP(LP(dest+4), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ if (sz==16) { \ - t1 = LP(ref+12); \ - OP(LP(dest+8), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ - t0 = LP(ref+16); \ - OP(LP(dest+12), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ + t1 = LP(ref+12); \ + OP(LP(dest+8), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ + t0 = LP(ref+16); \ + OP(LP(dest+12), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ } \ - ref+=stride; \ - dest+= stride; \ - } while(--height); \ + ref+=stride; \ + dest+= stride; \ + } while(--height); \ } /* aligned */ -#define OP_Y0(sz,avg2) \ +#define OP_Y0(sz,avg2) \ { \ - uint32_t t0,t1,t2,t3,t; \ + uint32_t t0,t1,t2,t3,t; \ \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ + t0 = LP(ref+0); \ + t1 = LP(ref+4); \ if (sz==16) { \ - t2 = LP(ref+8); \ - t3 = LP(ref+12); \ + t2 = LP(ref+8); \ + t3 = LP(ref+12); \ } \ - do { \ - ref += stride; \ + do { \ + ref += stride; \ \ - t = LP(ref+0); \ - OP(LP(dest+0), avg2(t0,t)); t0 = t; \ - t = LP(ref+4); \ - OP(LP(dest+4), avg2(t1,t)); t1 = t; \ + t = LP(ref+0); \ + OP(LP(dest+0), avg2(t0,t)); t0 = t; \ + t = LP(ref+4); \ + OP(LP(dest+4), avg2(t1,t)); t1 = t; \ if (sz==16) { \ - t = LP(ref+8); \ - OP(LP(dest+8), avg2(t2,t)); t2 = t; \ - t = LP(ref+12); \ - OP(LP(dest+12), avg2(t3,t)); t3 = t; \ + t = LP(ref+8); \ + OP(LP(dest+8), avg2(t2,t)); t2 = t; \ + t = LP(ref+12); \ + OP(LP(dest+12), avg2(t3,t)); t3 = t; \ } \ - dest+= stride; \ - } while(--height); \ + dest+= stride; \ + } while(--height); \ } -#define OP_Y(ofs,sz,avg2) \ +#define OP_Y(ofs,sz,avg2) \ { \ - uint32_t t0,t1,t2,t3,t,w0,w1; \ + uint32_t t0,t1,t2,t3,t,w0,w1; \ \ - ref-=ofs; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ - t0 = MERGE1(w0,w1,ofs); \ - w0 = LP(ref+8); \ - t1 = MERGE1(w1,w0,ofs); \ + ref-=ofs; \ + w0 = LP(ref+0); \ + w1 = LP(ref+4); \ + t0 = MERGE1(w0,w1,ofs); \ + w0 = LP(ref+8); \ + t1 = MERGE1(w1,w0,ofs); \ if (sz==16) { \ - w1 = LP(ref+12); \ - t2 = MERGE1(w0,w1,ofs); \ - w0 = LP(ref+16); \ - t3 = MERGE1(w1,w0,ofs); \ + w1 = LP(ref+12); \ + t2 = MERGE1(w0,w1,ofs); \ + w0 = LP(ref+16); \ + t3 = MERGE1(w1,w0,ofs); \ } \ - do { \ - ref += stride; \ + do { \ + ref += stride; \ \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ - t = MERGE1(w0,w1,ofs); \ - OP(LP(dest+0), avg2(t0,t)); t0 = t; \ - w0 = LP(ref+8); \ - t = MERGE1(w1,w0,ofs); \ - OP(LP(dest+4), avg2(t1,t)); t1 = t; \ + w0 = LP(ref+0); \ + w1 = LP(ref+4); \ + t = MERGE1(w0,w1,ofs); \ + OP(LP(dest+0), avg2(t0,t)); t0 = t; \ + w0 = LP(ref+8); \ + t = MERGE1(w1,w0,ofs); \ + OP(LP(dest+4), avg2(t1,t)); t1 = t; \ if (sz==16) { \ - w1 = LP(ref+12); \ - t = MERGE1(w0,w1,ofs); \ - OP(LP(dest+8), avg2(t2,t)); t2 = t; \ - w0 = LP(ref+16); \ - t = MERGE1(w1,w0,ofs); \ - OP(LP(dest+12), avg2(t3,t)); t3 = t; \ + w1 = LP(ref+12); \ + t = MERGE1(w0,w1,ofs); \ + OP(LP(dest+8), avg2(t2,t)); t2 = t; \ + w0 = LP(ref+16); \ + t = MERGE1(w1,w0,ofs); \ + OP(LP(dest+12), avg2(t3,t)); t3 = t; \ } \ - dest+=stride; \ - } while(--height); \ + dest+=stride; \ + } while(--height); \ } #define OP_X0(sz,avg2) OP_X(0,sz,avg2) #define OP_XY0(sz,PACK) OP_XY(0,sz,PACK) -#define OP_XY(ofs,sz,PACK) \ +#define OP_XY(ofs,sz,PACK) \ { \ - uint32_t t2,t3,w0,w1; \ - uint32_t a0,a1,a2,a3,a4,a5,a6,a7; \ + uint32_t t2,t3,w0,w1; \ + uint32_t a0,a1,a2,a3,a4,a5,a6,a7; \ \ - ref -= ofs; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ - UNPACK(a0,a1,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - w0 = LP(ref+8); \ - UNPACK(a2,a3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ + ref -= ofs; \ + w0 = LP(ref+0); \ + w1 = LP(ref+4); \ + UNPACK(a0,a1,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ + w0 = LP(ref+8); \ + UNPACK(a2,a3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ if (sz==16) { \ - w1 = LP(ref+12); \ - UNPACK(a4,a5,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - w0 = LP(ref+16); \ - UNPACK(a6,a7,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ + w1 = LP(ref+12); \ + UNPACK(a4,a5,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ + w0 = LP(ref+16); \ + UNPACK(a6,a7,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ } \ - do { \ - ref+=stride; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ - UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - OP(LP(dest+0),PACK(a0,a1,t2,t3)); \ - a0 = t2; a1 = t3; \ - w0 = LP(ref+8); \ - UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ - OP(LP(dest+4),PACK(a2,a3,t2,t3)); \ - a2 = t2; a3 = t3; \ + do { \ + ref+=stride; \ + w0 = LP(ref+0); \ + w1 = LP(ref+4); \ + UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ + OP(LP(dest+0),PACK(a0,a1,t2,t3)); \ + a0 = t2; a1 = t3; \ + w0 = LP(ref+8); \ + UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ + OP(LP(dest+4),PACK(a2,a3,t2,t3)); \ + a2 = t2; a3 = t3; \ if (sz==16) { \ - w1 = LP(ref+12); \ - UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - OP(LP(dest+8),PACK(a4,a5,t2,t3)); \ - a4 = t2; a5 = t3; \ - w0 = LP(ref+16); \ - UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ - OP(LP(dest+12),PACK(a6,a7,t2,t3)); \ - a6 = t2; a7 = t3; \ + w1 = LP(ref+12); \ + UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ + OP(LP(dest+8),PACK(a4,a5,t2,t3)); \ + a4 = t2; a5 = t3; \ + w0 = LP(ref+16); \ + UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ + OP(LP(dest+12),PACK(a6,a7,t2,t3)); \ + a6 = t2; a7 = t3; \ } \ - dest+=stride; \ - } while(--height); \ + dest+=stride; \ + } while(--height); \ } -#define DEFFUNC(op,rnd,xy,sz,OP_N,avgfunc) \ -static void op##_##rnd##_pixels##sz##_##xy (uint8_t * dest, const uint8_t * ref, \ - const int stride, int height) \ +#define DEFFUNC(op,rnd,xy,sz,OP_N,avgfunc) \ +static void op##_##rnd##_pixels##sz##_##xy (uint8_t * dest, const uint8_t * ref, \ + const int stride, int height) \ { \ - switch((int)ref&3) { \ - case 0:OP_N##0(sz,rnd##_##avgfunc); return; \ - case 1:OP_N(1,sz,rnd##_##avgfunc); return; \ - case 2:OP_N(2,sz,rnd##_##avgfunc); return; \ - case 3:OP_N(3,sz,rnd##_##avgfunc); return; \ - } \ + switch((int)ref&3) { \ + case 0:OP_N##0(sz,rnd##_##avgfunc); return; \ + case 1:OP_N(1,sz,rnd##_##avgfunc); return; \ + case 2:OP_N(2,sz,rnd##_##avgfunc); return; \ + case 3:OP_N(3,sz,rnd##_##avgfunc); return; \ + } \ } #define OP put @@ -305,21 +305,21 @@ #undef OP -#define put_no_rnd_pixels8_o put_rnd_pixels8_o -#define put_no_rnd_pixels16_o put_rnd_pixels16_o -#define avg_no_rnd_pixels8_o avg_rnd_pixels8_o -#define avg_no_rnd_pixels16_o avg_rnd_pixels16_o +#define put_no_rnd_pixels8_o put_rnd_pixels8_o +#define put_no_rnd_pixels16_o put_rnd_pixels16_o +#define avg_no_rnd_pixels8_o avg_rnd_pixels8_o +#define avg_no_rnd_pixels16_o avg_rnd_pixels16_o -#define put_pixels8_c put_rnd_pixels8_o -#define put_pixels16_c put_rnd_pixels16_o -#define avg_pixels8_c avg_rnd_pixels8_o -#define avg_pixels16_c avg_rnd_pixels16_o -#define put_no_rnd_pixels8_c put_rnd_pixels8_o -#define put_no_rnd_pixels16_c put_rnd_pixels16_o -#define avg_no_rnd_pixels8_c avg_rnd_pixels8_o -#define avg_no_rnd_pixels16_c avg_rnd_pixels16_o +#define put_pixels8_c put_rnd_pixels8_o +#define put_pixels16_c put_rnd_pixels16_o +#define avg_pixels8_c avg_rnd_pixels8_o +#define avg_pixels16_c avg_rnd_pixels16_o +#define put_no_rnd_pixels8_c put_rnd_pixels8_o +#define put_no_rnd_pixels16_c put_rnd_pixels16_o +#define avg_no_rnd_pixels8_c avg_rnd_pixels8_o +#define avg_no_rnd_pixels16_c avg_rnd_pixels16_o -#define QPEL +#define QPEL #ifdef QPEL @@ -329,41 +329,41 @@ void dsputil_init_align(DSPContext* c, AVCodecContext *avctx) { - c->put_pixels_tab[0][0] = put_rnd_pixels16_o; - c->put_pixels_tab[0][1] = put_rnd_pixels16_x; - c->put_pixels_tab[0][2] = put_rnd_pixels16_y; - c->put_pixels_tab[0][3] = put_rnd_pixels16_xy; - c->put_pixels_tab[1][0] = put_rnd_pixels8_o; - c->put_pixels_tab[1][1] = put_rnd_pixels8_x; - c->put_pixels_tab[1][2] = put_rnd_pixels8_y; - c->put_pixels_tab[1][3] = put_rnd_pixels8_xy; + c->put_pixels_tab[0][0] = put_rnd_pixels16_o; + c->put_pixels_tab[0][1] = put_rnd_pixels16_x; + c->put_pixels_tab[0][2] = put_rnd_pixels16_y; + c->put_pixels_tab[0][3] = put_rnd_pixels16_xy; + c->put_pixels_tab[1][0] = put_rnd_pixels8_o; + c->put_pixels_tab[1][1] = put_rnd_pixels8_x; + c->put_pixels_tab[1][2] = put_rnd_pixels8_y; + c->put_pixels_tab[1][3] = put_rnd_pixels8_xy; - c->put_no_rnd_pixels_tab[0][0] = put_no_rnd_pixels16_o; - c->put_no_rnd_pixels_tab[0][1] = put_no_rnd_pixels16_x; - c->put_no_rnd_pixels_tab[0][2] = put_no_rnd_pixels16_y; - c->put_no_rnd_pixels_tab[0][3] = put_no_rnd_pixels16_xy; - c->put_no_rnd_pixels_tab[1][0] = put_no_rnd_pixels8_o; - c->put_no_rnd_pixels_tab[1][1] = put_no_rnd_pixels8_x; - c->put_no_rnd_pixels_tab[1][2] = put_no_rnd_pixels8_y; - c->put_no_rnd_pixels_tab[1][3] = put_no_rnd_pixels8_xy; + c->put_no_rnd_pixels_tab[0][0] = put_no_rnd_pixels16_o; + c->put_no_rnd_pixels_tab[0][1] = put_no_rnd_pixels16_x; + c->put_no_rnd_pixels_tab[0][2] = put_no_rnd_pixels16_y; + c->put_no_rnd_pixels_tab[0][3] = put_no_rnd_pixels16_xy; + c->put_no_rnd_pixels_tab[1][0] = put_no_rnd_pixels8_o; + c->put_no_rnd_pixels_tab[1][1] = put_no_rnd_pixels8_x; + c->put_no_rnd_pixels_tab[1][2] = put_no_rnd_pixels8_y; + c->put_no_rnd_pixels_tab[1][3] = put_no_rnd_pixels8_xy; - c->avg_pixels_tab[0][0] = avg_rnd_pixels16_o; - c->avg_pixels_tab[0][1] = avg_rnd_pixels16_x; - c->avg_pixels_tab[0][2] = avg_rnd_pixels16_y; - c->avg_pixels_tab[0][3] = avg_rnd_pixels16_xy; - c->avg_pixels_tab[1][0] = avg_rnd_pixels8_o; - c->avg_pixels_tab[1][1] = avg_rnd_pixels8_x; - c->avg_pixels_tab[1][2] = avg_rnd_pixels8_y; - c->avg_pixels_tab[1][3] = avg_rnd_pixels8_xy; + c->avg_pixels_tab[0][0] = avg_rnd_pixels16_o; + c->avg_pixels_tab[0][1] = avg_rnd_pixels16_x; + c->avg_pixels_tab[0][2] = avg_rnd_pixels16_y; + c->avg_pixels_tab[0][3] = avg_rnd_pixels16_xy; + c->avg_pixels_tab[1][0] = avg_rnd_pixels8_o; + c->avg_pixels_tab[1][1] = avg_rnd_pixels8_x; + c->avg_pixels_tab[1][2] = avg_rnd_pixels8_y; + c->avg_pixels_tab[1][3] = avg_rnd_pixels8_xy; - c->avg_no_rnd_pixels_tab[0][0] = avg_no_rnd_pixels16_o; - c->avg_no_rnd_pixels_tab[0][1] = avg_no_rnd_pixels16_x; - c->avg_no_rnd_pixels_tab[0][2] = avg_no_rnd_pixels16_y; - c->avg_no_rnd_pixels_tab[0][3] = avg_no_rnd_pixels16_xy; - c->avg_no_rnd_pixels_tab[1][0] = avg_no_rnd_pixels8_o; - c->avg_no_rnd_pixels_tab[1][1] = avg_no_rnd_pixels8_x; - c->avg_no_rnd_pixels_tab[1][2] = avg_no_rnd_pixels8_y; - c->avg_no_rnd_pixels_tab[1][3] = avg_no_rnd_pixels8_xy; + c->avg_no_rnd_pixels_tab[0][0] = avg_no_rnd_pixels16_o; + c->avg_no_rnd_pixels_tab[0][1] = avg_no_rnd_pixels16_x; + c->avg_no_rnd_pixels_tab[0][2] = avg_no_rnd_pixels16_y; + c->avg_no_rnd_pixels_tab[0][3] = avg_no_rnd_pixels16_xy; + c->avg_no_rnd_pixels_tab[1][0] = avg_no_rnd_pixels8_o; + c->avg_no_rnd_pixels_tab[1][1] = avg_no_rnd_pixels8_x; + c->avg_no_rnd_pixels_tab[1][2] = avg_no_rnd_pixels8_y; + c->avg_no_rnd_pixels_tab[1][3] = avg_no_rnd_pixels8_xy; #ifdef QPEL
--- a/sh4/dsputil_sh4.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sh4/dsputil_sh4.c Thu Dec 22 01:10:11 2005 +0000 @@ -24,95 +24,95 @@ static void memzero_align8(void *dst,size_t size) { #if defined(__SH4__) || defined(__SH4_SINGLE__) || defined(__SH4_SINGLE_ONLY__) - (char*)dst+=size; - size/=8*4; - asm( + (char*)dst+=size; + size/=8*4; + asm( #if defined(__SH4__) - " fschg\n" //single float mode + " fschg\n" //single float mode #endif - " fldi0 fr0\n" - " fldi0 fr1\n" - " fschg\n" // double - "1: \n" \ - " dt %1\n" - " fmov dr0,@-%0\n" - " fmov dr0,@-%0\n" - " fmov dr0,@-%0\n" - " bf.s 1b\n" - " fmov dr0,@-%0\n" + " fldi0 fr0\n" + " fldi0 fr1\n" + " fschg\n" // double + "1: \n" \ + " dt %1\n" + " fmov dr0,@-%0\n" + " fmov dr0,@-%0\n" + " fmov dr0,@-%0\n" + " bf.s 1b\n" + " fmov dr0,@-%0\n" #if defined(__SH4_SINGLE__) || defined(__SH4_SINGLE_ONLY__) - " fschg" //back to single + " fschg" //back to single #endif - : : "r"(dst),"r"(size): "memory" ); + : : "r"(dst),"r"(size): "memory" ); #else - double *d = dst; - size/=8*4; - do { - d[0] = 0.0; - d[1] = 0.0; - d[2] = 0.0; - d[3] = 0.0; - d+=4; - } while(--size); + double *d = dst; + size/=8*4; + do { + d[0] = 0.0; + d[1] = 0.0; + d[2] = 0.0; + d[3] = 0.0; + d+=4; + } while(--size); #endif } static void clear_blocks_sh4(DCTELEM *blocks) { -// if (((int)blocks&7)==0) - memzero_align8(blocks,sizeof(DCTELEM)*6*64); +// if (((int)blocks&7)==0) + memzero_align8(blocks,sizeof(DCTELEM)*6*64); } extern void idct_sh4(DCTELEM *block); static void idct_put(uint8_t *dest, int line_size, DCTELEM *block) { - idct_sh4(block); - int i; - uint8_t *cm = cropTbl + MAX_NEG_CROP; - for(i=0;i<8;i++) { - dest[0] = cm[block[0]]; - dest[1] = cm[block[1]]; - dest[2] = cm[block[2]]; - dest[3] = cm[block[3]]; - dest[4] = cm[block[4]]; - dest[5] = cm[block[5]]; - dest[6] = cm[block[6]]; - dest[7] = cm[block[7]]; - dest+=line_size; - block+=8; - } + idct_sh4(block); + int i; + uint8_t *cm = cropTbl + MAX_NEG_CROP; + for(i=0;i<8;i++) { + dest[0] = cm[block[0]]; + dest[1] = cm[block[1]]; + dest[2] = cm[block[2]]; + dest[3] = cm[block[3]]; + dest[4] = cm[block[4]]; + dest[5] = cm[block[5]]; + dest[6] = cm[block[6]]; + dest[7] = cm[block[7]]; + dest+=line_size; + block+=8; + } } static void idct_add(uint8_t *dest, int line_size, DCTELEM *block) { - idct_sh4(block); - int i; - uint8_t *cm = cropTbl + MAX_NEG_CROP; - for(i=0;i<8;i++) { - dest[0] = cm[dest[0]+block[0]]; - dest[1] = cm[dest[1]+block[1]]; - dest[2] = cm[dest[2]+block[2]]; - dest[3] = cm[dest[3]+block[3]]; - dest[4] = cm[dest[4]+block[4]]; - dest[5] = cm[dest[5]+block[5]]; - dest[6] = cm[dest[6]+block[6]]; - dest[7] = cm[dest[7]+block[7]]; - dest+=line_size; - block+=8; - } + idct_sh4(block); + int i; + uint8_t *cm = cropTbl + MAX_NEG_CROP; + for(i=0;i<8;i++) { + dest[0] = cm[dest[0]+block[0]]; + dest[1] = cm[dest[1]+block[1]]; + dest[2] = cm[dest[2]+block[2]]; + dest[3] = cm[dest[3]+block[3]]; + dest[4] = cm[dest[4]+block[4]]; + dest[5] = cm[dest[5]+block[5]]; + dest[6] = cm[dest[6]+block[6]]; + dest[7] = cm[dest[7]+block[7]]; + dest+=line_size; + block+=8; + } } extern void dsputil_init_align(DSPContext* c, AVCodecContext *avctx); void dsputil_init_sh4(DSPContext* c, AVCodecContext *avctx) { - const int idct_algo= avctx->idct_algo; - dsputil_init_align(c,avctx); + const int idct_algo= avctx->idct_algo; + dsputil_init_align(c,avctx); - c->clear_blocks = clear_blocks_sh4; - if(idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_SH4){ - c->idct_put = idct_put; - c->idct_add = idct_add; + c->clear_blocks = clear_blocks_sh4; + if(idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_SH4){ + c->idct_put = idct_put; + c->idct_add = idct_add; c->idct = idct_sh4; - c->idct_permutation_type= FF_NO_IDCT_PERM; //FF_SIMPLE_IDCT_PERM; //FF_LIBMPEG2_IDCT_PERM; - } + c->idct_permutation_type= FF_NO_IDCT_PERM; //FF_SIMPLE_IDCT_PERM; //FF_LIBMPEG2_IDCT_PERM; + } }
--- a/sh4/idct_sh4.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sh4/idct_sh4.c Thu Dec 22 01:10:11 2005 +0000 @@ -19,65 +19,65 @@ */ #include "../dsputil.h" -#define c1 1.38703984532214752434 /* sqrt(2)*cos(1*pi/16) */ -#define c2 1.30656296487637657577 /* sqrt(2)*cos(2*pi/16) */ -#define c3 1.17587560241935884520 /* sqrt(2)*cos(3*pi/16) */ -#define c4 1.00000000000000000000 /* sqrt(2)*cos(4*pi/16) */ -#define c5 0.78569495838710234903 /* sqrt(2)*cos(5*pi/16) */ -#define c6 0.54119610014619712324 /* sqrt(2)*cos(6*pi/16) */ -#define c7 0.27589937928294311353 /* sqrt(2)*cos(7*pi/16) */ +#define c1 1.38703984532214752434 /* sqrt(2)*cos(1*pi/16) */ +#define c2 1.30656296487637657577 /* sqrt(2)*cos(2*pi/16) */ +#define c3 1.17587560241935884520 /* sqrt(2)*cos(3*pi/16) */ +#define c4 1.00000000000000000000 /* sqrt(2)*cos(4*pi/16) */ +#define c5 0.78569495838710234903 /* sqrt(2)*cos(5*pi/16) */ +#define c6 0.54119610014619712324 /* sqrt(2)*cos(6*pi/16) */ +#define c7 0.27589937928294311353 /* sqrt(2)*cos(7*pi/16) */ -const static float even_table[] __attribute__ ((aligned(8))) = { - c4, c4, c4, c4, - c2, c6,-c6,-c2, - c4,-c4,-c4, c4, - c6,-c2, c2,-c6 +const static float even_table[] __attribute__ ((aligned(8))) = { + c4, c4, c4, c4, + c2, c6,-c6,-c2, + c4,-c4,-c4, c4, + c6,-c2, c2,-c6 }; -const static float odd_table[] __attribute__ ((aligned(8))) = { - c1, c3, c5, c7, - c3,-c7,-c1,-c5, - c5,-c1, c7, c3, - c7,-c5, c3,-c1 +const static float odd_table[] __attribute__ ((aligned(8))) = { + c1, c3, c5, c7, + c3,-c7,-c1,-c5, + c5,-c1, c7, c3, + c7,-c5, c3,-c1 }; -#undef c1 -#undef c2 -#undef c3 -#undef c4 -#undef c5 -#undef c6 -#undef c7 +#undef c1 +#undef c2 +#undef c3 +#undef c4 +#undef c5 +#undef c6 +#undef c7 #if defined(__SH4_SINGLE__) || defined(__SH4_SINGLE_ONLY__) -#define load_matrix(table) \ - __asm__ volatile( \ - " fschg\n" \ - " fmov @%0+,xd0\n" \ - " fmov @%0+,xd2\n" \ - " fmov @%0+,xd4\n" \ - " fmov @%0+,xd6\n" \ - " fmov @%0+,xd8\n" \ - " fmov @%0+,xd10\n" \ - " fmov @%0+,xd12\n" \ - " fmov @%0+,xd14\n" \ - " fschg\n" \ - :\ - : "r"(table)\ - : "0" \ - ) +#define load_matrix(table) \ + __asm__ volatile( \ + " fschg\n" \ + " fmov @%0+,xd0\n" \ + " fmov @%0+,xd2\n" \ + " fmov @%0+,xd4\n" \ + " fmov @%0+,xd6\n" \ + " fmov @%0+,xd8\n" \ + " fmov @%0+,xd10\n" \ + " fmov @%0+,xd12\n" \ + " fmov @%0+,xd14\n" \ + " fschg\n" \ + :\ + : "r"(table)\ + : "0" \ + ) -#define ftrv() \ - __asm__ volatile("ftrv xmtrx,fv0" \ - : "=f"(fr0),"=f"(fr1),"=f"(fr2),"=f"(fr3) \ - : "0"(fr0), "1"(fr1), "2"(fr2), "3"(fr3) ); +#define ftrv() \ + __asm__ volatile("ftrv xmtrx,fv0" \ + : "=f"(fr0),"=f"(fr1),"=f"(fr2),"=f"(fr3) \ + : "0"(fr0), "1"(fr1), "2"(fr2), "3"(fr3) ); -#define DEFREG \ - register float fr0 __asm__("fr0"); \ - register float fr1 __asm__("fr1"); \ - register float fr2 __asm__("fr2"); \ - register float fr3 __asm__("fr3") +#define DEFREG \ + register float fr0 __asm__("fr0"); \ + register float fr1 __asm__("fr1"); \ + register float fr2 __asm__("fr2"); \ + register float fr3 __asm__("fr3") #else @@ -85,40 +85,40 @@ static void ftrv_(const float xf[],float fv[]) { - float f0,f1,f2,f3; - f0 = fv[0]; - f1 = fv[1]; - f2 = fv[2]; - f3 = fv[3]; - fv[0] = xf[0]*f0 + xf[4]*f1 + xf[ 8]*f2 + xf[12]*f3; - fv[1] = xf[1]*f0 + xf[5]*f1 + xf[ 9]*f2 + xf[13]*f3; - fv[2] = xf[2]*f0 + xf[6]*f1 + xf[10]*f2 + xf[14]*f3; - fv[3] = xf[3]*f0 + xf[7]*f1 + xf[11]*f2 + xf[15]*f3; + float f0,f1,f2,f3; + f0 = fv[0]; + f1 = fv[1]; + f2 = fv[2]; + f3 = fv[3]; + fv[0] = xf[0]*f0 + xf[4]*f1 + xf[ 8]*f2 + xf[12]*f3; + fv[1] = xf[1]*f0 + xf[5]*f1 + xf[ 9]*f2 + xf[13]*f3; + fv[2] = xf[2]*f0 + xf[6]*f1 + xf[10]*f2 + xf[14]*f3; + fv[3] = xf[3]*f0 + xf[7]*f1 + xf[11]*f2 + xf[15]*f3; } static void load_matrix_(float xf[],const float table[]) { - int i; - for(i=0;i<16;i++) xf[i]=table[i]; + int i; + for(i=0;i<16;i++) xf[i]=table[i]; } -#define ftrv() ftrv_(xf,fv) -#define load_matrix(table) load_matrix_(xf,table) +#define ftrv() ftrv_(xf,fv) +#define load_matrix(table) load_matrix_(xf,table) -#define DEFREG \ - float fv[4],xf[16] +#define DEFREG \ + float fv[4],xf[16] -#define fr0 fv[0] -#define fr1 fv[1] -#define fr2 fv[2] -#define fr3 fv[3] +#define fr0 fv[0] +#define fr1 fv[1] +#define fr2 fv[2] +#define fr3 fv[3] #endif #if 1 -#define DESCALE(x,n) (x)*(1.0f/(1<<(n))) +#define DESCALE(x,n) (x)*(1.0f/(1<<(n))) #else -#define DESCALE(x,n) (((int)(x)+(1<<(n-1)))>>(n)) +#define DESCALE(x,n) (((int)(x)+(1<<(n-1)))>>(n)) #endif /* this code work worse on gcc cvs. 3.2.3 work fine */ @@ -129,236 +129,236 @@ void idct_sh4(DCTELEM *block) { - DEFREG; + DEFREG; - int i; - float tblock[8*8],*fblock; - int ofs1,ofs2,ofs3; + int i; + float tblock[8*8],*fblock; + int ofs1,ofs2,ofs3; #if defined(__SH4__) -#error "FIXME!! change to single float" +#error "FIXME!! change to single float" #endif - /* row */ + /* row */ - /* even part */ - load_matrix(even_table); + /* even part */ + load_matrix(even_table); - fblock = tblock+4; - i = 8; - do { - fr0 = block[0]; - fr1 = block[2]; - fr2 = block[4]; - fr3 = block[6]; - block+=8; - ftrv(); - *--fblock = fr3; - *--fblock = fr2; - *--fblock = fr1; - *--fblock = fr0; - fblock+=8+4; - } while(--i); - block-=8*8; - fblock-=8*8+4; + fblock = tblock+4; + i = 8; + do { + fr0 = block[0]; + fr1 = block[2]; + fr2 = block[4]; + fr3 = block[6]; + block+=8; + ftrv(); + *--fblock = fr3; + *--fblock = fr2; + *--fblock = fr1; + *--fblock = fr0; + fblock+=8+4; + } while(--i); + block-=8*8; + fblock-=8*8+4; - load_matrix(odd_table); + load_matrix(odd_table); - i = 8; + i = 8; -// ofs1 = sizeof(float)*1; -// ofs2 = sizeof(float)*2; -// ofs3 = sizeof(float)*3; +// ofs1 = sizeof(float)*1; +// ofs2 = sizeof(float)*2; +// ofs3 = sizeof(float)*3; - do { - float t0,t1,t2,t3; - fr0 = block[1]; - fr1 = block[3]; - fr2 = block[5]; - fr3 = block[7]; - block+=8; - ftrv(); - t0 = *fblock++; - t1 = *fblock++; - t2 = *fblock++; - t3 = *fblock++; - fblock+=4; - *--fblock = t0 - fr0; - *--fblock = t1 - fr1; - *--fblock = t2 - fr2; - *--fblock = t3 - fr3; - *--fblock = t3 + fr3; - *--fblock = t2 + fr2; - *--fblock = t1 + fr1; - *--fblock = t0 + fr0; - fblock+=8; - } while(--i); - block-=8*8; - fblock-=8*8; + do { + float t0,t1,t2,t3; + fr0 = block[1]; + fr1 = block[3]; + fr2 = block[5]; + fr3 = block[7]; + block+=8; + ftrv(); + t0 = *fblock++; + t1 = *fblock++; + t2 = *fblock++; + t3 = *fblock++; + fblock+=4; + *--fblock = t0 - fr0; + *--fblock = t1 - fr1; + *--fblock = t2 - fr2; + *--fblock = t3 - fr3; + *--fblock = t3 + fr3; + *--fblock = t2 + fr2; + *--fblock = t1 + fr1; + *--fblock = t0 + fr0; + fblock+=8; + } while(--i); + block-=8*8; + fblock-=8*8; - /* col */ + /* col */ - /* even part */ - load_matrix(even_table); + /* even part */ + load_matrix(even_table); - ofs1 = sizeof(float)*2*8; - ofs2 = sizeof(float)*4*8; - ofs3 = sizeof(float)*6*8; + ofs1 = sizeof(float)*2*8; + ofs2 = sizeof(float)*4*8; + ofs3 = sizeof(float)*6*8; - i = 8; + i = 8; -#define OA(fblock,ofs) *(float*)((char*)fblock + ofs) +#define OA(fblock,ofs) *(float*)((char*)fblock + ofs) - do { - fr0 = OA(fblock, 0); - fr1 = OA(fblock,ofs1); - fr2 = OA(fblock,ofs2); - fr3 = OA(fblock,ofs3); - ftrv(); - OA(fblock,0 ) = fr0; - OA(fblock,ofs1) = fr1; - OA(fblock,ofs2) = fr2; - OA(fblock,ofs3) = fr3; - fblock++; - } while(--i); - fblock-=8; + do { + fr0 = OA(fblock, 0); + fr1 = OA(fblock,ofs1); + fr2 = OA(fblock,ofs2); + fr3 = OA(fblock,ofs3); + ftrv(); + OA(fblock,0 ) = fr0; + OA(fblock,ofs1) = fr1; + OA(fblock,ofs2) = fr2; + OA(fblock,ofs3) = fr3; + fblock++; + } while(--i); + fblock-=8; - load_matrix(odd_table); + load_matrix(odd_table); - i=8; - do { - float t0,t1,t2,t3; - t0 = OA(fblock, 0); /* [8*0] */ - t1 = OA(fblock,ofs1); /* [8*2] */ - t2 = OA(fblock,ofs2); /* [8*4] */ - t3 = OA(fblock,ofs3); /* [8*6] */ - fblock+=8; - fr0 = OA(fblock, 0); /* [8*1] */ - fr1 = OA(fblock,ofs1); /* [8*3] */ - fr2 = OA(fblock,ofs2); /* [8*5] */ - fr3 = OA(fblock,ofs3); /* [8*7] */ - fblock+=-8+1; - ftrv(); - block[8*0] = DESCALE(t0 + fr0,3); - block[8*7] = DESCALE(t0 - fr0,3); - block[8*1] = DESCALE(t1 + fr1,3); - block[8*6] = DESCALE(t1 - fr1,3); - block[8*2] = DESCALE(t2 + fr2,3); - block[8*5] = DESCALE(t2 - fr2,3); - block[8*3] = DESCALE(t3 + fr3,3); - block[8*4] = DESCALE(t3 - fr3,3); - block++; - } while(--i); + i=8; + do { + float t0,t1,t2,t3; + t0 = OA(fblock, 0); /* [8*0] */ + t1 = OA(fblock,ofs1); /* [8*2] */ + t2 = OA(fblock,ofs2); /* [8*4] */ + t3 = OA(fblock,ofs3); /* [8*6] */ + fblock+=8; + fr0 = OA(fblock, 0); /* [8*1] */ + fr1 = OA(fblock,ofs1); /* [8*3] */ + fr2 = OA(fblock,ofs2); /* [8*5] */ + fr3 = OA(fblock,ofs3); /* [8*7] */ + fblock+=-8+1; + ftrv(); + block[8*0] = DESCALE(t0 + fr0,3); + block[8*7] = DESCALE(t0 - fr0,3); + block[8*1] = DESCALE(t1 + fr1,3); + block[8*6] = DESCALE(t1 - fr1,3); + block[8*2] = DESCALE(t2 + fr2,3); + block[8*5] = DESCALE(t2 - fr2,3); + block[8*3] = DESCALE(t3 + fr3,3); + block[8*4] = DESCALE(t3 - fr3,3); + block++; + } while(--i); #if defined(__SH4__) -#error "FIXME!! change to double" +#error "FIXME!! change to double" #endif } #else void idct_sh4(DCTELEM *block) { - DEFREG; + DEFREG; - int i; - float tblock[8*8],*fblock; + int i; + float tblock[8*8],*fblock; - /* row */ + /* row */ - /* even part */ - load_matrix(even_table); + /* even part */ + load_matrix(even_table); - fblock = tblock; - i = 8; - do { - fr0 = block[0]; - fr1 = block[2]; - fr2 = block[4]; - fr3 = block[6]; - block+=8; - ftrv(); - fblock[0] = fr0; - fblock[2] = fr1; - fblock[4] = fr2; - fblock[6] = fr3; - fblock+=8; - } while(--i); - block-=8*8; - fblock-=8*8; + fblock = tblock; + i = 8; + do { + fr0 = block[0]; + fr1 = block[2]; + fr2 = block[4]; + fr3 = block[6]; + block+=8; + ftrv(); + fblock[0] = fr0; + fblock[2] = fr1; + fblock[4] = fr2; + fblock[6] = fr3; + fblock+=8; + } while(--i); + block-=8*8; + fblock-=8*8; - load_matrix(odd_table); + load_matrix(odd_table); - i = 8; + i = 8; - do { - float t0,t1,t2,t3; - fr0 = block[1]; - fr1 = block[3]; - fr2 = block[5]; - fr3 = block[7]; - block+=8; - ftrv(); - t0 = fblock[0]; - t1 = fblock[2]; - t2 = fblock[4]; - t3 = fblock[6]; - fblock[0] = t0 + fr0; - fblock[7] = t0 - fr0; - fblock[1] = t1 + fr1; - fblock[6] = t1 - fr1; - fblock[2] = t2 + fr2; - fblock[5] = t2 - fr2; - fblock[3] = t3 + fr3; - fblock[4] = t3 - fr3; - fblock+=8; - } while(--i); - block-=8*8; - fblock-=8*8; + do { + float t0,t1,t2,t3; + fr0 = block[1]; + fr1 = block[3]; + fr2 = block[5]; + fr3 = block[7]; + block+=8; + ftrv(); + t0 = fblock[0]; + t1 = fblock[2]; + t2 = fblock[4]; + t3 = fblock[6]; + fblock[0] = t0 + fr0; + fblock[7] = t0 - fr0; + fblock[1] = t1 + fr1; + fblock[6] = t1 - fr1; + fblock[2] = t2 + fr2; + fblock[5] = t2 - fr2; + fblock[3] = t3 + fr3; + fblock[4] = t3 - fr3; + fblock+=8; + } while(--i); + block-=8*8; + fblock-=8*8; - /* col */ + /* col */ - /* even part */ - load_matrix(even_table); + /* even part */ + load_matrix(even_table); - i = 8; + i = 8; - do { - fr0 = fblock[8*0]; - fr1 = fblock[8*2]; - fr2 = fblock[8*4]; - fr3 = fblock[8*6]; - ftrv(); - fblock[8*0] = fr0; - fblock[8*2] = fr1; - fblock[8*4] = fr2; - fblock[8*6] = fr3; - fblock++; - } while(--i); - fblock-=8; + do { + fr0 = fblock[8*0]; + fr1 = fblock[8*2]; + fr2 = fblock[8*4]; + fr3 = fblock[8*6]; + ftrv(); + fblock[8*0] = fr0; + fblock[8*2] = fr1; + fblock[8*4] = fr2; + fblock[8*6] = fr3; + fblock++; + } while(--i); + fblock-=8; - load_matrix(odd_table); + load_matrix(odd_table); - i=8; - do { - float t0,t1,t2,t3; - fr0 = fblock[8*1]; - fr1 = fblock[8*3]; - fr2 = fblock[8*5]; - fr3 = fblock[8*7]; - ftrv(); - t0 = fblock[8*0]; - t1 = fblock[8*2]; - t2 = fblock[8*4]; - t3 = fblock[8*6]; - fblock++; - block[8*0] = DESCALE(t0 + fr0,3); - block[8*7] = DESCALE(t0 - fr0,3); - block[8*1] = DESCALE(t1 + fr1,3); - block[8*6] = DESCALE(t1 - fr1,3); - block[8*2] = DESCALE(t2 + fr2,3); - block[8*5] = DESCALE(t2 - fr2,3); - block[8*3] = DESCALE(t3 + fr3,3); - block[8*4] = DESCALE(t3 - fr3,3); - block++; - } while(--i); + i=8; + do { + float t0,t1,t2,t3; + fr0 = fblock[8*1]; + fr1 = fblock[8*3]; + fr2 = fblock[8*5]; + fr3 = fblock[8*7]; + ftrv(); + t0 = fblock[8*0]; + t1 = fblock[8*2]; + t2 = fblock[8*4]; + t3 = fblock[8*6]; + fblock++; + block[8*0] = DESCALE(t0 + fr0,3); + block[8*7] = DESCALE(t0 - fr0,3); + block[8*1] = DESCALE(t1 + fr1,3); + block[8*6] = DESCALE(t1 - fr1,3); + block[8*2] = DESCALE(t2 + fr2,3); + block[8*5] = DESCALE(t2 - fr2,3); + block[8*3] = DESCALE(t3 + fr3,3); + block[8*4] = DESCALE(t3 - fr3,3); + block++; + } while(--i); } #endif
--- a/sh4/qpel.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sh4/qpel.c Thu Dec 22 01:10:11 2005 +0000 @@ -1,183 +1,183 @@ /* - this is optimized for sh, which have post increment addressing (*p++) - some cpu may be index (p[n]) faster than post increment (*p++) + this is optimized for sh, which have post increment addressing (*p++) + some cpu may be index (p[n]) faster than post increment (*p++) */ -#define LD(adr) *(uint32_t*)(adr) +#define LD(adr) *(uint32_t*)(adr) #define PIXOP2(OPNAME, OP) \ /*static inline void OPNAME ## _no_rnd_pixels8_l2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels8_l2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels4_l2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels16_l2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ - OP(LP(dst+8),no_rnd_avg32(LD32(src1+8),LD32(src2+8)) ); \ - OP(LP(dst+12),no_rnd_avg32(LD32(src1+12),LD32(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ + OP(LP(dst+8),no_rnd_avg32(LD32(src1+8),LD32(src2+8)) ); \ + OP(LP(dst+12),no_rnd_avg32(LD32(src1+12),LD32(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels16_l2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ - OP(LP(dst+8),rnd_avg32(LD32(src1+8),LD32(src2+8)) ); \ - OP(LP(dst+12),rnd_avg32(LD32(src1+12),LD32(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LD32(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LD32(src1+4),LD32(src2+4)) ); \ + OP(LP(dst+8),rnd_avg32(LD32(src1+8),LD32(src2+8)) ); \ + OP(LP(dst+12),rnd_avg32(LD32(src1+12),LD32(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }*/\ \ static inline void OPNAME ## _pixels4_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels4_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels16_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),no_rnd_avg32(LD32(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),no_rnd_avg32(LD32(src1+12),LP(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ + OP(LP(dst+8),no_rnd_avg32(LD32(src1+8),LP(src2+8)) ); \ + OP(LP(dst+12),no_rnd_avg32(LD32(src1+12),LP(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels16_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),rnd_avg32(LD32(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),rnd_avg32(LD32(src1+12),LP(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ + OP(LP(dst+8),rnd_avg32(LD32(src1+8),LP(src2+8)) ); \ + OP(LP(dst+12),rnd_avg32(LD32(src1+12),LP(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels8_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do { /* onlye src2 aligned */\ - OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do { /* onlye src2 aligned */\ + OP(LP(dst ),no_rnd_avg32(LD32(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels8_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LD32(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LD32(src1+4),LP(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels8_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels8_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels16_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),no_rnd_avg32(LP(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),no_rnd_avg32(LP(src1+12),LP(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + OP(LP(dst+8),no_rnd_avg32(LP(src1+8),LP(src2+8)) ); \ + OP(LP(dst+12),no_rnd_avg32(LP(src1+12),LP(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _pixels16_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ - do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),rnd_avg32(LP(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),rnd_avg32(LP(src1+12),LP(src2+12)) ); \ - src1+=src_stride1; \ - src2+=src_stride2; \ - dst+=dst_stride; \ - } while(--h); \ + do {\ + OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + OP(LP(dst+8),rnd_avg32(LP(src1+8),LP(src2+8)) ); \ + OP(LP(dst+12),rnd_avg32(LP(src1+12),LP(src2+12)) ); \ + src1+=src_stride1; \ + src2+=src_stride2; \ + dst+=dst_stride; \ + } while(--h); \ }\ \ static inline void OPNAME ## _no_rnd_pixels16_l2_aligned1(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ @@ -193,163 +193,163 @@ { OPNAME ## _pixels8_l2_aligned2(dst,src2,src1,dst_stride,src_stride2,src_stride1,h); } \ \ static inline void OPNAME ## _pixels8_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LP(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _no_rnd_pixels8_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LP(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _pixels8_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; /* src1 only not aligned */\ - UNPACK(a0,a1,LD32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; /* src1 only not aligned */\ + UNPACK(a0,a1,LD32(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _no_rnd_pixels8_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LD32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LD32(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _pixels16_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ - OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ - OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LP(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ + UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ + UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _no_rnd_pixels16_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ - OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ - OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LP(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ + UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ + UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _pixels16_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { /* src1 is unaligned */\ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LD32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ - OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ - OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { /* src1 is unaligned */\ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LD32(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+8),LP(src2+8)); \ + UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+12),LP(src2+12)); \ + UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \ static inline void OPNAME ## _no_rnd_pixels16_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ - do { \ - uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LD32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ - OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ - OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ - OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LD32(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ - OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ - src1+=src_stride1;\ - src2+=src_stride2;\ - src3+=src_stride3;\ - src4+=src_stride4;\ - dst+=dst_stride;\ - } while(--h); \ + do { \ + uint32_t a0,a1,a2,a3; \ + UNPACK(a0,a1,LD32(src1),LP(src2)); \ + UNPACK(a2,a3,LP(src3),LP(src4)); \ + OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+4),LP(src2+4)); \ + UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+8),LP(src2+8)); \ + UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ + UNPACK(a0,a1,LD32(src1+12),LP(src2+12)); \ + UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ + src1+=src_stride1;\ + src2+=src_stride2;\ + src3+=src_stride3;\ + src4+=src_stride4;\ + dst+=dst_stride;\ + } while(--h); \ } \ \
--- a/simple_idct.c Wed Dec 21 17:50:40 2005 +0000 +++ b/simple_idct.c Thu Dec 22 01:10:11 2005 +0000 @@ -75,7 +75,7 @@ static inline void idctRowCondDC (DCTELEM * row) { - int a0, a1, a2, a3, b0, b1, b2, b3; + int a0, a1, a2, a3, b0, b1, b2, b3; #ifdef FAST_64BIT uint64_t temp; #else @@ -97,7 +97,7 @@ ((uint64_t *)row)[0] = temp; ((uint64_t *)row)[1] = temp; return; - } + } }else{ if (!(row[1]|row[2]|row[3]|row[4]|row[5]|row[6]|row[7])) { row[0]=row[1]=row[2]=row[3]=row[4]=row[5]=row[6]=row[7]= row[0] << 3; @@ -125,9 +125,9 @@ #endif a0 = (W4 * row[0]) + (1 << (ROW_SHIFT - 1)); - a1 = a0; - a2 = a0; - a3 = a0; + a1 = a0; + a2 = a0; + a3 = a0; /* no need to optimize : gcc does it */ a0 += W2 * row[2]; @@ -149,7 +149,7 @@ #else temp = ((uint32_t*)row)[2] | ((uint32_t*)row)[3]; #endif - if (temp != 0) { + if (temp != 0) { a0 += W4*row[4] + W6*row[6]; a1 += - W4*row[4] - W2*row[6]; a2 += - W4*row[4] + W2*row[6]; @@ -166,29 +166,29 @@ MAC16(b3, W3, row[5]); MAC16(b3, -W1, row[7]); - } + } - row[0] = (a0 + b0) >> ROW_SHIFT; - row[7] = (a0 - b0) >> ROW_SHIFT; - row[1] = (a1 + b1) >> ROW_SHIFT; - row[6] = (a1 - b1) >> ROW_SHIFT; - row[2] = (a2 + b2) >> ROW_SHIFT; - row[5] = (a2 - b2) >> ROW_SHIFT; - row[3] = (a3 + b3) >> ROW_SHIFT; - row[4] = (a3 - b3) >> ROW_SHIFT; + row[0] = (a0 + b0) >> ROW_SHIFT; + row[7] = (a0 - b0) >> ROW_SHIFT; + row[1] = (a1 + b1) >> ROW_SHIFT; + row[6] = (a1 - b1) >> ROW_SHIFT; + row[2] = (a2 + b2) >> ROW_SHIFT; + row[5] = (a2 - b2) >> ROW_SHIFT; + row[3] = (a3 + b3) >> ROW_SHIFT; + row[4] = (a3 - b3) >> ROW_SHIFT; } static inline void idctSparseColPut (uint8_t *dest, int line_size, DCTELEM * col) { - int a0, a1, a2, a3, b0, b1, b2, b3; + int a0, a1, a2, a3, b0, b1, b2, b3; uint8_t *cm = cropTbl + MAX_NEG_CROP; /* XXX: I did that only to give same values as previous code */ - a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); - a1 = a0; - a2 = a0; - a3 = a0; + a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); + a1 = a0; + a2 = a0; + a3 = a0; a0 += + W2*col[8*2]; a1 += + W6*col[8*2]; @@ -205,33 +205,33 @@ MAC16(b2, - W1, col[8*3]); MAC16(b3, - W5, col[8*3]); - if(col[8*4]){ + if(col[8*4]){ a0 += + W4*col[8*4]; a1 += - W4*col[8*4]; a2 += - W4*col[8*4]; a3 += + W4*col[8*4]; - } + } - if (col[8*5]) { + if (col[8*5]) { MAC16(b0, + W5, col[8*5]); MAC16(b1, - W1, col[8*5]); MAC16(b2, + W7, col[8*5]); MAC16(b3, + W3, col[8*5]); - } + } - if(col[8*6]){ + if(col[8*6]){ a0 += + W6*col[8*6]; a1 += - W2*col[8*6]; a2 += + W2*col[8*6]; a3 += - W6*col[8*6]; - } + } - if (col[8*7]) { + if (col[8*7]) { MAC16(b0, + W7, col[8*7]); MAC16(b1, - W5, col[8*7]); MAC16(b2, + W3, col[8*7]); MAC16(b3, - W1, col[8*7]); - } + } dest[0] = cm[(a0 + b0) >> COL_SHIFT]; dest += line_size; @@ -253,14 +253,14 @@ static inline void idctSparseColAdd (uint8_t *dest, int line_size, DCTELEM * col) { - int a0, a1, a2, a3, b0, b1, b2, b3; + int a0, a1, a2, a3, b0, b1, b2, b3; uint8_t *cm = cropTbl + MAX_NEG_CROP; /* XXX: I did that only to give same values as previous code */ - a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); - a1 = a0; - a2 = a0; - a3 = a0; + a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); + a1 = a0; + a2 = a0; + a3 = a0; a0 += + W2*col[8*2]; a1 += + W6*col[8*2]; @@ -277,33 +277,33 @@ MAC16(b2, - W1, col[8*3]); MAC16(b3, - W5, col[8*3]); - if(col[8*4]){ + if(col[8*4]){ a0 += + W4*col[8*4]; a1 += - W4*col[8*4]; a2 += - W4*col[8*4]; a3 += + W4*col[8*4]; - } + } - if (col[8*5]) { + if (col[8*5]) { MAC16(b0, + W5, col[8*5]); MAC16(b1, - W1, col[8*5]); MAC16(b2, + W7, col[8*5]); MAC16(b3, + W3, col[8*5]); - } + } - if(col[8*6]){ + if(col[8*6]){ a0 += + W6*col[8*6]; a1 += - W2*col[8*6]; a2 += + W2*col[8*6]; a3 += - W6*col[8*6]; - } + } - if (col[8*7]) { + if (col[8*7]) { MAC16(b0, + W7, col[8*7]); MAC16(b1, - W5, col[8*7]); MAC16(b2, + W3, col[8*7]); MAC16(b3, - W1, col[8*7]); - } + } dest[0] = cm[dest[0] + ((a0 + b0) >> COL_SHIFT)]; dest += line_size; @@ -324,13 +324,13 @@ static inline void idctSparseCol (DCTELEM * col) { - int a0, a1, a2, a3, b0, b1, b2, b3; + int a0, a1, a2, a3, b0, b1, b2, b3; /* XXX: I did that only to give same values as previous code */ - a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); - a1 = a0; - a2 = a0; - a3 = a0; + a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); + a1 = a0; + a2 = a0; + a3 = a0; a0 += + W2*col[8*2]; a1 += + W6*col[8*2]; @@ -347,33 +347,33 @@ MAC16(b2, - W1, col[8*3]); MAC16(b3, - W5, col[8*3]); - if(col[8*4]){ + if(col[8*4]){ a0 += + W4*col[8*4]; a1 += - W4*col[8*4]; a2 += - W4*col[8*4]; a3 += + W4*col[8*4]; - } + } - if (col[8*5]) { + if (col[8*5]) { MAC16(b0, + W5, col[8*5]); MAC16(b1, - W1, col[8*5]); MAC16(b2, + W7, col[8*5]); MAC16(b3, + W3, col[8*5]); - } + } - if(col[8*6]){ + if(col[8*6]){ a0 += + W6*col[8*6]; a1 += - W2*col[8*6]; a2 += + W2*col[8*6]; a3 += - W6*col[8*6]; - } + } - if (col[8*7]) { + if (col[8*7]) { MAC16(b0, + W7, col[8*7]); MAC16(b1, - W5, col[8*7]); MAC16(b2, + W3, col[8*7]); MAC16(b3, - W1, col[8*7]); - } + } col[0 ] = ((a0 + b0) >> COL_SHIFT); col[8 ] = ((a1 + b1) >> COL_SHIFT);
--- a/smc.c Wed Dec 21 17:50:40 2005 +0000 +++ b/smc.c Thu Dec 22 01:10:11 2005 +0000 @@ -195,7 +195,7 @@ /* sanity check */ if ((row_ptr == 0) && (pixel_ptr < 2 * 4)) { - av_log(s->avctx, AV_LOG_INFO, "encountered repeat block opcode (%02X) but not enough blocks rendered yet\n", + av_log(s->avctx, AV_LOG_INFO, "encountered repeat block opcode (%02X) but not enough blocks rendered yet\n", opcode & 0xF0); break; }
--- a/snow.c Wed Dec 21 17:50:40 2005 +0000 +++ b/snow.c Thu Dec 22 01:10:11 2005 +0000 @@ -566,9 +566,9 @@ av_free(buf->line); } -#ifdef __sgi +#ifdef __sgi // Avoid a name clash on SGI IRIX -#undef qexp +#undef qexp #endif #define QEXPSHIFT (7-FRAC_BITS+8) //FIXME try to change this to 0 static uint8_t qexp[QROOT];
--- a/sonic.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sonic.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,20 +64,20 @@ int *predictor_state[MAX_CHANNELS]; } SonicContext; -#define LATTICE_SHIFT 10 -#define SAMPLE_SHIFT 4 -#define LATTICE_FACTOR (1 << LATTICE_SHIFT) -#define SAMPLE_FACTOR (1 << SAMPLE_SHIFT) +#define LATTICE_SHIFT 10 +#define SAMPLE_SHIFT 4 +#define LATTICE_FACTOR (1 << LATTICE_SHIFT) +#define SAMPLE_FACTOR (1 << SAMPLE_SHIFT) -#define BASE_QUANT 0.6 -#define RATE_VARIATION 3.0 +#define BASE_QUANT 0.6 +#define RATE_VARIATION 3.0 static inline int divide(int a, int b) { if (a < 0) - return -( (-a + b/2)/b ); + return -( (-a + b/2)/b ); else - return (a + b/2)/b; + return (a + b/2)/b; } static inline int shift(int a,int b) @@ -96,7 +96,7 @@ int i; for (i = 0; i < entries; i++) - set_se_golomb(pb, buf[i]); + set_se_golomb(pb, buf[i]); return 1; } @@ -106,7 +106,7 @@ int i; for (i = 0; i < entries; i++) - buf[i] = get_se_golomb(gb); + buf[i] = get_se_golomb(gb); return 1; } @@ -121,8 +121,8 @@ while(x) { - res++; - x >>= 1; + res++; + x >>= 1; } return res; } @@ -132,15 +132,15 @@ int i, bits; if (!max) - return; + return; bits = bits_to_store(max); for (i = 0; i < bits-1; i++) - put_bits(pb, 1, value & (1 << i)); + put_bits(pb, 1, value & (1 << i)); if ( (value | (1 << (bits-1))) <= max) - put_bits(pb, 1, value & (1 << (bits-1))); + put_bits(pb, 1, value & (1 << (bits-1))); } static unsigned int read_uint_max(GetBitContext *gb, int max) @@ -148,17 +148,17 @@ int i, bits, value = 0; if (!max) - return 0; + return 0; bits = bits_to_store(max); for (i = 0; i < bits-1; i++) - if (get_bits1(gb)) - value += 1 << i; + if (get_bits1(gb)) + value += 1 << i; if ( (value | (1<<(bits-1))) <= max) - if (get_bits1(gb)) - value += 1 << (bits-1); + if (get_bits1(gb)) + value += 1 << (bits-1); return value; } @@ -171,88 +171,88 @@ copy = av_mallocz(4* entries); if (!copy) - return -1; + return -1; if (base_2_part) { - int energy = 0; + int energy = 0; - for (i = 0; i < entries; i++) - energy += abs(buf[i]); + for (i = 0; i < entries; i++) + energy += abs(buf[i]); - low_bits = bits_to_store(energy / (entries * 2)); - if (low_bits > 15) - low_bits = 15; + low_bits = bits_to_store(energy / (entries * 2)); + if (low_bits > 15) + low_bits = 15; - put_bits(pb, 4, low_bits); + put_bits(pb, 4, low_bits); } for (i = 0; i < entries; i++) { - put_bits(pb, low_bits, abs(buf[i])); - copy[i] = abs(buf[i]) >> low_bits; - if (copy[i] > max) - max = abs(copy[i]); + put_bits(pb, low_bits, abs(buf[i])); + copy[i] = abs(buf[i]) >> low_bits; + if (copy[i] > max) + max = abs(copy[i]); } bits = av_mallocz(4* entries*max); if (!bits) { -// av_free(copy); - return -1; +// av_free(copy); + return -1; } for (i = 0; i <= max; i++) { - for (j = 0; j < entries; j++) - if (copy[j] >= i) - bits[x++] = copy[j] > i; + for (j = 0; j < entries; j++) + if (copy[j] >= i) + bits[x++] = copy[j] > i; } // store bitstream while (pos < x) { - int steplet = step >> 8; + int steplet = step >> 8; - if (pos + steplet > x) - steplet = x - pos; + if (pos + steplet > x) + steplet = x - pos; - for (i = 0; i < steplet; i++) - if (bits[i+pos] != dominant) - any = 1; + for (i = 0; i < steplet; i++) + if (bits[i+pos] != dominant) + any = 1; - put_bits(pb, 1, any); + put_bits(pb, 1, any); - if (!any) - { - pos += steplet; - step += step / ADAPT_LEVEL; - } - else - { - int interloper = 0; + if (!any) + { + pos += steplet; + step += step / ADAPT_LEVEL; + } + else + { + int interloper = 0; - while (((pos + interloper) < x) && (bits[pos + interloper] == dominant)) - interloper++; + while (((pos + interloper) < x) && (bits[pos + interloper] == dominant)) + interloper++; - // note change - write_uint_max(pb, interloper, (step >> 8) - 1); + // note change + write_uint_max(pb, interloper, (step >> 8) - 1); - pos += interloper + 1; - step -= step / ADAPT_LEVEL; - } + pos += interloper + 1; + step -= step / ADAPT_LEVEL; + } - if (step < 256) - { - step = 65536 / step; - dominant = !dominant; - } + if (step < 256) + { + step = 65536 / step; + dominant = !dominant; + } } // store signs for (i = 0; i < entries; i++) - if (buf[i]) - put_bits(pb, 1, buf[i] < 0); + if (buf[i]) + put_bits(pb, 1, buf[i] < 0); // av_free(bits); // av_free(copy); @@ -268,90 +268,90 @@ int *bits = av_mallocz(4* entries); if (!bits) - return -1; + return -1; if (base_2_part) { - low_bits = get_bits(gb, 4); + low_bits = get_bits(gb, 4); - if (low_bits) - for (i = 0; i < entries; i++) - buf[i] = get_bits(gb, low_bits); + if (low_bits) + for (i = 0; i < entries; i++) + buf[i] = get_bits(gb, low_bits); } // av_log(NULL, AV_LOG_INFO, "entries: %d, low bits: %d\n", entries, low_bits); while (n_zeros < entries) { - int steplet = step >> 8; + int steplet = step >> 8; - if (!get_bits1(gb)) - { - for (i = 0; i < steplet; i++) - bits[x++] = dominant; + if (!get_bits1(gb)) + { + for (i = 0; i < steplet; i++) + bits[x++] = dominant; - if (!dominant) - n_zeros += steplet; + if (!dominant) + n_zeros += steplet; - step += step / ADAPT_LEVEL; - } - else - { - int actual_run = read_uint_max(gb, steplet-1); + step += step / ADAPT_LEVEL; + } + else + { + int actual_run = read_uint_max(gb, steplet-1); -// av_log(NULL, AV_LOG_INFO, "actual run: %d\n", actual_run); +// av_log(NULL, AV_LOG_INFO, "actual run: %d\n", actual_run); - for (i = 0; i < actual_run; i++) - bits[x++] = dominant; + for (i = 0; i < actual_run; i++) + bits[x++] = dominant; - bits[x++] = !dominant; + bits[x++] = !dominant; - if (!dominant) - n_zeros += actual_run; - else - n_zeros++; + if (!dominant) + n_zeros += actual_run; + else + n_zeros++; - step -= step / ADAPT_LEVEL; - } + step -= step / ADAPT_LEVEL; + } - if (step < 256) - { - step = 65536 / step; - dominant = !dominant; - } + if (step < 256) + { + step = 65536 / step; + dominant = !dominant; + } } // reconstruct unsigned values n_zeros = 0; for (i = 0; n_zeros < entries; i++) { - while(1) - { - if (pos >= entries) - { - pos = 0; - level += 1 << low_bits; - } + while(1) + { + if (pos >= entries) + { + pos = 0; + level += 1 << low_bits; + } - if (buf[pos] >= level) - break; + if (buf[pos] >= level) + break; - pos++; - } + pos++; + } - if (bits[i]) - buf[pos] += 1 << low_bits; - else - n_zeros++; + if (bits[i]) + buf[pos] += 1 << low_bits; + else + n_zeros++; - pos++; + pos++; } // av_free(bits); // read signs for (i = 0; i < entries; i++) - if (buf[i] && get_bits1(gb)) - buf[i] = -buf[i]; + if (buf[i] && get_bits1(gb)) + buf[i] = -buf[i]; // av_log(NULL, AV_LOG_INFO, "zeros: %d pos: %d\n", n_zeros, pos); @@ -365,14 +365,14 @@ for (i = order-2; i >= 0; i--) { - int j, p, x = state[i]; + int j, p, x = state[i]; - for (j = 0, p = i+1; p < order; j++,p++) - { - int tmp = x + shift_down(k[j] * state[p], LATTICE_SHIFT); - state[p] += shift_down(k[j]*x, LATTICE_SHIFT); - x = tmp; - } + for (j = 0, p = i+1; p < order; j++,p++) + { + int tmp = x + shift_down(k[j] * state[p], LATTICE_SHIFT); + state[p] += shift_down(k[j]*x, LATTICE_SHIFT); + x = tmp; + } } } @@ -382,18 +382,18 @@ #if 1 int *k_ptr = &(k[order-2]), - *state_ptr = &(state[order-2]); + *state_ptr = &(state[order-2]); for (i = order-2; i >= 0; i--, k_ptr--, state_ptr--) { - int k_value = *k_ptr, state_value = *state_ptr; - x -= shift_down(k_value * state_value, LATTICE_SHIFT); - state_ptr[1] = state_value + shift_down(k_value * x, LATTICE_SHIFT); + int k_value = *k_ptr, state_value = *state_ptr; + x -= shift_down(k_value * state_value, LATTICE_SHIFT); + state_ptr[1] = state_value + shift_down(k_value * x, LATTICE_SHIFT); } #else for (i = order-2; i >= 0; i--) { - x -= shift_down(k[i] * state[i], LATTICE_SHIFT); - state[i+1] = state[i] + shift_down(k[i] * x, LATTICE_SHIFT); + x -= shift_down(k[i] * state[i], LATTICE_SHIFT); + state[i+1] = state[i] + shift_down(k[i] * x, LATTICE_SHIFT); } #endif @@ -411,7 +411,7 @@ // actual whitened result as it goes. static void modified_levinson_durbin(int *window, int window_entries, - int *out, int out_entries, int channels, int *tap_quant) + int *out, int out_entries, int channels, int *tap_quant) { int i; int *state = av_mallocz(4* window_entries); @@ -420,57 +420,57 @@ for (i = 0; i < out_entries; i++) { - int step = (i+1)*channels, k, j; - double xx = 0.0, xy = 0.0; + int step = (i+1)*channels, k, j; + double xx = 0.0, xy = 0.0; #if 1 - int *x_ptr = &(window[step]), *state_ptr = &(state[0]); - j = window_entries - step; - for (;j>=0;j--,x_ptr++,state_ptr++) - { - double x_value = *x_ptr, state_value = *state_ptr; - xx += state_value*state_value; - xy += x_value*state_value; - } + int *x_ptr = &(window[step]), *state_ptr = &(state[0]); + j = window_entries - step; + for (;j>=0;j--,x_ptr++,state_ptr++) + { + double x_value = *x_ptr, state_value = *state_ptr; + xx += state_value*state_value; + xy += x_value*state_value; + } #else - for (j = 0; j <= (window_entries - step); j++); - { - double stepval = window[step+j], stateval = window[j]; -// xx += (double)window[j]*(double)window[j]; -// xy += (double)window[step+j]*(double)window[j]; - xx += stateval*stateval; - xy += stepval*stateval; - } + for (j = 0; j <= (window_entries - step); j++); + { + double stepval = window[step+j], stateval = window[j]; +// xx += (double)window[j]*(double)window[j]; +// xy += (double)window[step+j]*(double)window[j]; + xx += stateval*stateval; + xy += stepval*stateval; + } #endif - if (xx == 0.0) - k = 0; - else - k = (int)(floor(-xy/xx * (double)LATTICE_FACTOR / (double)(tap_quant[i]) + 0.5)); + if (xx == 0.0) + k = 0; + else + k = (int)(floor(-xy/xx * (double)LATTICE_FACTOR / (double)(tap_quant[i]) + 0.5)); - if (k > (LATTICE_FACTOR/tap_quant[i])) - k = LATTICE_FACTOR/tap_quant[i]; - if (-k > (LATTICE_FACTOR/tap_quant[i])) - k = -(LATTICE_FACTOR/tap_quant[i]); + if (k > (LATTICE_FACTOR/tap_quant[i])) + k = LATTICE_FACTOR/tap_quant[i]; + if (-k > (LATTICE_FACTOR/tap_quant[i])) + k = -(LATTICE_FACTOR/tap_quant[i]); - out[i] = k; - k *= tap_quant[i]; + out[i] = k; + k *= tap_quant[i]; #if 1 - x_ptr = &(window[step]); - state_ptr = &(state[0]); - j = window_entries - step; - for (;j>=0;j--,x_ptr++,state_ptr++) - { - int x_value = *x_ptr, state_value = *state_ptr; - *x_ptr = x_value + shift_down(k*state_value,LATTICE_SHIFT); - *state_ptr = state_value + shift_down(k*x_value, LATTICE_SHIFT); - } + x_ptr = &(window[step]); + state_ptr = &(state[0]); + j = window_entries - step; + for (;j>=0;j--,x_ptr++,state_ptr++) + { + int x_value = *x_ptr, state_value = *state_ptr; + *x_ptr = x_value + shift_down(k*state_value,LATTICE_SHIFT); + *state_ptr = state_value + shift_down(k*x_value, LATTICE_SHIFT); + } #else - for (j=0; j <= (window_entries - step); j++) - { - int stepval = window[step+j], stateval=state[j]; - window[step+j] += shift_down(k * stateval, LATTICE_SHIFT); - state[j] += shift_down(k * stepval, LATTICE_SHIFT); - } + for (j=0; j <= (window_entries - step); j++) + { + int stepval = window[step+j], stateval=state[j]; + window[step+j] += shift_down(k * stateval, LATTICE_SHIFT); + state[j] += shift_down(k * stepval, LATTICE_SHIFT); + } #endif } @@ -486,15 +486,15 @@ { switch (samplerate) { - case 44100: return 0; - case 22050: return 1; - case 11025: return 2; - case 96000: return 3; - case 48000: return 4; - case 32000: return 5; - case 24000: return 6; - case 16000: return 7; - case 8000: return 8; + case 44100: return 0; + case 22050: return 1; + case 11025: return 2; + case 96000: return 3; + case 48000: return 4; + case 32000: return 5; + case 24000: return 6; + case 16000: return 7; + case 8000: return 8; } return -1; } @@ -507,39 +507,39 @@ if (avctx->channels > MAX_CHANNELS) { - av_log(avctx, AV_LOG_ERROR, "Only mono and stereo streams are supported by now\n"); + av_log(avctx, AV_LOG_ERROR, "Only mono and stereo streams are supported by now\n"); return -1; /* only stereo or mono for now */ } if (avctx->channels == 2) - s->decorrelation = MID_SIDE; + s->decorrelation = MID_SIDE; if (avctx->codec->id == CODEC_ID_SONIC_LS) { - s->lossless = 1; - s->num_taps = 32; - s->downsampling = 1; - s->quantization = 0.0; + s->lossless = 1; + s->num_taps = 32; + s->downsampling = 1; + s->quantization = 0.0; } else { - s->num_taps = 128; - s->downsampling = 2; - s->quantization = 1.0; + s->num_taps = 128; + s->downsampling = 2; + s->quantization = 1.0; } // max tap 2048 if ((s->num_taps < 32) || (s->num_taps > 1024) || - ((s->num_taps>>5)<<5 != s->num_taps)) + ((s->num_taps>>5)<<5 != s->num_taps)) { - av_log(avctx, AV_LOG_ERROR, "Invalid number of taps\n"); - return -1; + av_log(avctx, AV_LOG_ERROR, "Invalid number of taps\n"); + return -1; } // generate taps s->tap_quant = av_mallocz(4* s->num_taps); for (i = 0; i < s->num_taps; i++) - s->tap_quant[i] = (int)(sqrt(i+1)); + s->tap_quant[i] = (int)(sqrt(i+1)); s->channels = avctx->channels; s->samplerate = avctx->sample_rate; @@ -549,18 +549,18 @@ s->tail = av_mallocz(4* s->num_taps*s->channels); if (!s->tail) - return -1; + return -1; s->tail_size = s->num_taps*s->channels; s->predictor_k = av_mallocz(4 * s->num_taps); if (!s->predictor_k) - return -1; + return -1; for (i = 0; i < s->channels; i++) { - s->coded_samples[i] = av_mallocz(4* s->block_align); - if (!s->coded_samples[i]) - return -1; + s->coded_samples[i] = av_mallocz(4* s->block_align); + if (!s->coded_samples[i]) + return -1; } s->int_samples = av_mallocz(4* s->frame_size); @@ -568,22 +568,22 @@ s->window_size = ((2*s->tail_size)+s->frame_size); s->window = av_mallocz(4* s->window_size); if (!s->window) - return -1; + return -1; avctx->extradata = av_mallocz(16); if (!avctx->extradata) - return -1; + return -1; init_put_bits(&pb, avctx->extradata, 16*8); put_bits(&pb, 2, version); // version if (version == 1) { - put_bits(&pb, 2, s->channels); - put_bits(&pb, 4, code_samplerate(s->samplerate)); + put_bits(&pb, 2, s->channels); + put_bits(&pb, 4, code_samplerate(s->samplerate)); } put_bits(&pb, 1, s->lossless); if (!s->lossless) - put_bits(&pb, 3, SAMPLE_SHIFT); // XXX FIXME: sample precision + put_bits(&pb, 3, SAMPLE_SHIFT); // XXX FIXME: sample precision put_bits(&pb, 2, s->decorrelation); put_bits(&pb, 2, s->downsampling); put_bits(&pb, 5, (s->num_taps >> 5)-1); // 32..1024 @@ -593,11 +593,11 @@ avctx->extradata_size = put_bits_count(&pb)/8; av_log(avctx, AV_LOG_INFO, "Sonic: ver: %d ls: %d dr: %d taps: %d block: %d frame: %d downsamp: %d\n", - version, s->lossless, s->decorrelation, s->num_taps, s->block_align, s->frame_size, s->downsampling); + version, s->lossless, s->decorrelation, s->num_taps, s->block_align, s->frame_size, s->downsampling); avctx->coded_frame = avcodec_alloc_frame(); if (!avctx->coded_frame) - return -ENOMEM; + return -ENOMEM; avctx->coded_frame->key_frame = 1; avctx->frame_size = s->block_align*s->downsampling; @@ -612,7 +612,7 @@ av_freep(&avctx->coded_frame); for (i = 0; i < s->channels; i++) - av_free(s->coded_samples[i]); + av_free(s->coded_samples[i]); av_free(s->predictor_k); av_free(s->tail); @@ -624,7 +624,7 @@ } static int sonic_encode_frame(AVCodecContext *avctx, - uint8_t *buf, int buf_size, void *data) + uint8_t *buf, int buf_size, void *data) { SonicContext *s = avctx->priv_data; PutBitContext pb; @@ -635,108 +635,108 @@ // short -> internal for (i = 0; i < s->frame_size; i++) - s->int_samples[i] = samples[i]; + s->int_samples[i] = samples[i]; if (!s->lossless) - for (i = 0; i < s->frame_size; i++) - s->int_samples[i] = s->int_samples[i] << SAMPLE_SHIFT; + for (i = 0; i < s->frame_size; i++) + s->int_samples[i] = s->int_samples[i] << SAMPLE_SHIFT; switch(s->decorrelation) { - case MID_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - { - s->int_samples[i] += s->int_samples[i+1]; - s->int_samples[i+1] -= shift(s->int_samples[i], 1); - } - break; - case LEFT_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - s->int_samples[i+1] -= s->int_samples[i]; - break; - case RIGHT_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - s->int_samples[i] -= s->int_samples[i+1]; - break; + case MID_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + { + s->int_samples[i] += s->int_samples[i+1]; + s->int_samples[i+1] -= shift(s->int_samples[i], 1); + } + break; + case LEFT_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + s->int_samples[i+1] -= s->int_samples[i]; + break; + case RIGHT_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + s->int_samples[i] -= s->int_samples[i+1]; + break; } memset(s->window, 0, 4* s->window_size); for (i = 0; i < s->tail_size; i++) - s->window[x++] = s->tail[i]; + s->window[x++] = s->tail[i]; for (i = 0; i < s->frame_size; i++) - s->window[x++] = s->int_samples[i]; + s->window[x++] = s->int_samples[i]; for (i = 0; i < s->tail_size; i++) - s->window[x++] = 0; + s->window[x++] = 0; for (i = 0; i < s->tail_size; i++) - s->tail[i] = s->int_samples[s->frame_size - s->tail_size + i]; + s->tail[i] = s->int_samples[s->frame_size - s->tail_size + i]; // generate taps modified_levinson_durbin(s->window, s->window_size, - s->predictor_k, s->num_taps, s->channels, s->tap_quant); + s->predictor_k, s->num_taps, s->channels, s->tap_quant); if (intlist_write(&pb, s->predictor_k, s->num_taps, 0) < 0) - return -1; + return -1; for (ch = 0; ch < s->channels; ch++) { - x = s->tail_size+ch; - for (i = 0; i < s->block_align; i++) - { - int sum = 0; - for (j = 0; j < s->downsampling; j++, x += s->channels) - sum += s->window[x]; - s->coded_samples[ch][i] = sum; - } + x = s->tail_size+ch; + for (i = 0; i < s->block_align; i++) + { + int sum = 0; + for (j = 0; j < s->downsampling; j++, x += s->channels) + sum += s->window[x]; + s->coded_samples[ch][i] = sum; + } } // simple rate control code if (!s->lossless) { - double energy1 = 0.0, energy2 = 0.0; - for (ch = 0; ch < s->channels; ch++) - { - for (i = 0; i < s->block_align; i++) - { - double sample = s->coded_samples[ch][i]; - energy2 += sample*sample; - energy1 += fabs(sample); - } - } + double energy1 = 0.0, energy2 = 0.0; + for (ch = 0; ch < s->channels; ch++) + { + for (i = 0; i < s->block_align; i++) + { + double sample = s->coded_samples[ch][i]; + energy2 += sample*sample; + energy1 += fabs(sample); + } + } - energy2 = sqrt(energy2/(s->channels*s->block_align)); - energy1 = sqrt(2.0)*energy1/(s->channels*s->block_align); + energy2 = sqrt(energy2/(s->channels*s->block_align)); + energy1 = sqrt(2.0)*energy1/(s->channels*s->block_align); - // increase bitrate when samples are like a gaussian distribution - // reduce bitrate when samples are like a two-tailed exponential distribution + // increase bitrate when samples are like a gaussian distribution + // reduce bitrate when samples are like a two-tailed exponential distribution - if (energy2 > energy1) - energy2 += (energy2-energy1)*RATE_VARIATION; + if (energy2 > energy1) + energy2 += (energy2-energy1)*RATE_VARIATION; - quant = (int)(BASE_QUANT*s->quantization*energy2/SAMPLE_FACTOR); -// av_log(avctx, AV_LOG_DEBUG, "quant: %d energy: %f / %f\n", quant, energy1, energy2); + quant = (int)(BASE_QUANT*s->quantization*energy2/SAMPLE_FACTOR); +// av_log(avctx, AV_LOG_DEBUG, "quant: %d energy: %f / %f\n", quant, energy1, energy2); - if (quant < 1) - quant = 1; - if (quant > 65535) - quant = 65535; + if (quant < 1) + quant = 1; + if (quant > 65535) + quant = 65535; - set_ue_golomb(&pb, quant); + set_ue_golomb(&pb, quant); - quant *= SAMPLE_FACTOR; + quant *= SAMPLE_FACTOR; } // write out coded samples for (ch = 0; ch < s->channels; ch++) { - if (!s->lossless) - for (i = 0; i < s->block_align; i++) - s->coded_samples[ch][i] = divide(s->coded_samples[ch][i], quant); + if (!s->lossless) + for (i = 0; i < s->block_align; i++) + s->coded_samples[ch][i] = divide(s->coded_samples[ch][i], quant); - if (intlist_write(&pb, s->coded_samples[ch], s->block_align, 1) < 0) - return -1; + if (intlist_write(&pb, s->coded_samples[ch], s->block_align, 1) < 0) + return -1; } // av_log(avctx, AV_LOG_DEBUG, "used bytes: %d\n", (put_bits_count(&pb)+7)/8); @@ -757,8 +757,8 @@ if (!avctx->extradata) { - av_log(avctx, AV_LOG_ERROR, "No mandatory headers present\n"); - return -1; + av_log(avctx, AV_LOG_ERROR, "No mandatory headers present\n"); + return -1; } init_get_bits(&gb, avctx->extradata, avctx->extradata_size); @@ -766,60 +766,60 @@ version = get_bits(&gb, 2); if (version > 1) { - av_log(avctx, AV_LOG_ERROR, "Unsupported Sonic version, please report\n"); - return -1; + av_log(avctx, AV_LOG_ERROR, "Unsupported Sonic version, please report\n"); + return -1; } if (version == 1) { - s->channels = get_bits(&gb, 2); - s->samplerate = samplerate_table[get_bits(&gb, 4)]; - av_log(avctx, AV_LOG_INFO, "Sonicv2 chans: %d samprate: %d\n", - s->channels, s->samplerate); + s->channels = get_bits(&gb, 2); + s->samplerate = samplerate_table[get_bits(&gb, 4)]; + av_log(avctx, AV_LOG_INFO, "Sonicv2 chans: %d samprate: %d\n", + s->channels, s->samplerate); } if (s->channels > MAX_CHANNELS) { - av_log(avctx, AV_LOG_ERROR, "Only mono and stereo streams are supported by now\n"); - return -1; + av_log(avctx, AV_LOG_ERROR, "Only mono and stereo streams are supported by now\n"); + return -1; } s->lossless = get_bits1(&gb); if (!s->lossless) - skip_bits(&gb, 3); // XXX FIXME + skip_bits(&gb, 3); // XXX FIXME s->decorrelation = get_bits(&gb, 2); s->downsampling = get_bits(&gb, 2); s->num_taps = (get_bits(&gb, 5)+1)<<5; if (get_bits1(&gb)) // XXX FIXME - av_log(avctx, AV_LOG_INFO, "Custom quant table\n"); + av_log(avctx, AV_LOG_INFO, "Custom quant table\n"); s->block_align = (int)(2048.0*(s->samplerate/44100))/s->downsampling; s->frame_size = s->channels*s->block_align*s->downsampling; // avctx->frame_size = s->block_align; av_log(avctx, AV_LOG_INFO, "Sonic: ver: %d ls: %d dr: %d taps: %d block: %d frame: %d downsamp: %d\n", - version, s->lossless, s->decorrelation, s->num_taps, s->block_align, s->frame_size, s->downsampling); + version, s->lossless, s->decorrelation, s->num_taps, s->block_align, s->frame_size, s->downsampling); // generate taps s->tap_quant = av_mallocz(4* s->num_taps); for (i = 0; i < s->num_taps; i++) - s->tap_quant[i] = (int)(sqrt(i+1)); + s->tap_quant[i] = (int)(sqrt(i+1)); s->predictor_k = av_mallocz(4* s->num_taps); for (i = 0; i < s->channels; i++) { - s->predictor_state[i] = av_mallocz(4* s->num_taps); - if (!s->predictor_state[i]) - return -1; + s->predictor_state[i] = av_mallocz(4* s->num_taps); + if (!s->predictor_state[i]) + return -1; } for (i = 0; i < s->channels; i++) { - s->coded_samples[i] = av_mallocz(4* s->block_align); - if (!s->coded_samples[i]) - return -1; + s->coded_samples[i] = av_mallocz(4* s->block_align); + if (!s->coded_samples[i]) + return -1; } s->int_samples = av_mallocz(4* s->frame_size); @@ -837,16 +837,16 @@ for (i = 0; i < s->channels; i++) { - av_free(s->predictor_state[i]); - av_free(s->coded_samples[i]); + av_free(s->predictor_state[i]); + av_free(s->coded_samples[i]); } return 0; } static int sonic_decode_frame(AVCodecContext *avctx, - void *data, int *data_size, - uint8_t *buf, int buf_size) + void *data, int *data_size, + uint8_t *buf, int buf_size) { SonicContext *s = avctx->priv_data; GetBitContext gb; @@ -863,71 +863,71 @@ // dequantize for (i = 0; i < s->num_taps; i++) - s->predictor_k[i] *= s->tap_quant[i]; + s->predictor_k[i] *= s->tap_quant[i]; if (s->lossless) - quant = 1; + quant = 1; else - quant = get_ue_golomb(&gb) * SAMPLE_FACTOR; + quant = get_ue_golomb(&gb) * SAMPLE_FACTOR; // av_log(NULL, AV_LOG_INFO, "quant: %d\n", quant); for (ch = 0; ch < s->channels; ch++) { - int x = ch; + int x = ch; - predictor_init_state(s->predictor_k, s->predictor_state[ch], s->num_taps); + predictor_init_state(s->predictor_k, s->predictor_state[ch], s->num_taps); - intlist_read(&gb, s->coded_samples[ch], s->block_align, 1); + intlist_read(&gb, s->coded_samples[ch], s->block_align, 1); - for (i = 0; i < s->block_align; i++) - { - for (j = 0; j < s->downsampling - 1; j++) - { - s->int_samples[x] = predictor_calc_error(s->predictor_k, s->predictor_state[ch], s->num_taps, 0); - x += s->channels; - } + for (i = 0; i < s->block_align; i++) + { + for (j = 0; j < s->downsampling - 1; j++) + { + s->int_samples[x] = predictor_calc_error(s->predictor_k, s->predictor_state[ch], s->num_taps, 0); + x += s->channels; + } - s->int_samples[x] = predictor_calc_error(s->predictor_k, s->predictor_state[ch], s->num_taps, s->coded_samples[ch][i] * quant); - x += s->channels; - } + s->int_samples[x] = predictor_calc_error(s->predictor_k, s->predictor_state[ch], s->num_taps, s->coded_samples[ch][i] * quant); + x += s->channels; + } - for (i = 0; i < s->num_taps; i++) - s->predictor_state[ch][i] = s->int_samples[s->frame_size - s->channels + ch - i*s->channels]; + for (i = 0; i < s->num_taps; i++) + s->predictor_state[ch][i] = s->int_samples[s->frame_size - s->channels + ch - i*s->channels]; } switch(s->decorrelation) { - case MID_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - { - s->int_samples[i+1] += shift(s->int_samples[i], 1); - s->int_samples[i] -= s->int_samples[i+1]; - } - break; - case LEFT_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - s->int_samples[i+1] += s->int_samples[i]; - break; - case RIGHT_SIDE: - for (i = 0; i < s->frame_size; i += s->channels) - s->int_samples[i] += s->int_samples[i+1]; - break; + case MID_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + { + s->int_samples[i+1] += shift(s->int_samples[i], 1); + s->int_samples[i] -= s->int_samples[i+1]; + } + break; + case LEFT_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + s->int_samples[i+1] += s->int_samples[i]; + break; + case RIGHT_SIDE: + for (i = 0; i < s->frame_size; i += s->channels) + s->int_samples[i] += s->int_samples[i+1]; + break; } if (!s->lossless) - for (i = 0; i < s->frame_size; i++) - s->int_samples[i] = shift(s->int_samples[i], SAMPLE_SHIFT); + for (i = 0; i < s->frame_size; i++) + s->int_samples[i] = shift(s->int_samples[i], SAMPLE_SHIFT); // internal -> short for (i = 0; i < s->frame_size; i++) { - if (s->int_samples[i] > 32767) - samples[i] = 32767; - else if (s->int_samples[i] < -32768) - samples[i] = -32768; - else - samples[i] = s->int_samples[i]; + if (s->int_samples[i] > 32767) + samples[i] = 32767; + else if (s->int_samples[i] < -32768) + samples[i] = -32768; + else + samples[i] = s->int_samples[i]; } align_get_bits(&gb);
--- a/sp5x.h Wed Dec 21 17:50:40 2005 +0000 +++ b/sp5x.h Thu Dec 22 01:10:11 2005 +0000 @@ -22,12 +22,12 @@ static const uint8_t sp5x_data_sof[] = { - 0xFF, 0xC0, /* SOF */ - 0x00, 0x11, /* len */ - 0x08, /* bits */ - 0x00, 0xf0, /* height (default: 240) */ - 0x01, 0x40, /* width (default: 240) */ - 0x03, /* nb components */ + 0xFF, 0xC0, /* SOF */ + 0x00, 0x11, /* len */ + 0x08, /* bits */ + 0x00, 0xf0, /* height (default: 240) */ + 0x01, 0x40, /* width (default: 240) */ + 0x03, /* nb components */ 0x01, 0x22, 0x00, /* 21 vs 22 ? */ 0x02, 0x11, 0x01, 0x03, 0x11, 0x01 @@ -35,15 +35,15 @@ static const uint8_t sp5x_data_sos[] = { - 0xFF, 0xDA, /* SOS */ - 0x00, 0x0C, /* len */ - 0x03, /* nb components */ + 0xFF, 0xDA, /* SOS */ + 0x00, 0x0C, /* len */ + 0x03, /* nb components */ 0x01, 0x00, 0x02, 0x11, 0x03, 0x11, - 0x00, /* Ss */ - 0x3F, /* Se */ - 0x00 /* Ah/Al */ + 0x00, /* Ss */ + 0x3F, /* Se */ + 0x00 /* Ah/Al */ }; static const uint8_t sp5x_data_dqt[] = @@ -71,8 +71,8 @@ }; static const uint8_t sp5x_data_dht[] = { - 0xFF, 0xC4, /* DHT */ - 0x01, 0xA2, /* len */ + 0xFF, 0xC4, /* DHT */ + 0x01, 0xA2, /* len */ 0x00, 0x00, 0x01, 0x05, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, @@ -257,7 +257,7 @@ /* index 2, Q80 */ { 6, 4, 4, 6, 10, 16, 20, 24, 5, 5, 6, 8, 10, 23, 24, 22, - 6, 5, 6, 10, 16, 23, 28, 22, 6, 7, 9, 12, 20, 35, 32, 25, + 6, 5, 6, 10, 16, 23, 28, 22, 6, 7, 9, 12, 20, 35, 32, 25, 7, 9, 15, 22, 27, 44, 41, 31, 10, 14, 22, 26, 32, 42, 45, 37, 20, 26, 31, 35, 41, 48, 48, 40, 29, 37, 38, 39, 45, 40, 41, 40 }, { 7, 7, 10, 19, 40, 40, 40, 40, 7, 8, 10, 26, 40, 40, 40, 40, @@ -267,7 +267,7 @@ /* index 3, Q85 */ { 5, 3, 3, 5, 7, 12, 15, 18, 4, 4, 4, 6, 8, 17, 18, 17, - 4, 4, 5, 7, 12, 17, 21, 17, 4, 5, 7, 9, 15, 26, 24, 19, + 4, 4, 5, 7, 12, 17, 21, 17, 4, 5, 7, 9, 15, 26, 24, 19, 5, 7, 11, 17, 20, 33, 31, 23, 7, 11, 17, 19, 24, 31, 34, 28, 15, 19, 23, 26, 31, 36, 36, 30, 22, 28, 29, 29, 34, 30, 31, 30 }, { 5, 5, 7, 14, 30, 30, 30, 30, 5, 6, 8, 20, 30, 30, 30, 30,
--- a/sparc/dsputil_vis.c Wed Dec 21 17:50:40 2005 +0000 +++ b/sparc/dsputil_vis.c Thu Dec 22 01:10:11 2005 +0000 @@ -48,12 +48,12 @@ * of '0xfe' is in f4, a repeating mask of '0x7f' is in f6, and * the value 0x80808080 is in f8): * - * fxor f0, f2, f10 - * fand f10, f4, f10 - * fmul8x16 f8, f10, f10 - * fand f10, f6, f10 - * for f0, f2, f12 - * fpsub16 f12, f10, f10 + * fxor f0, f2, f10 + * fand f10, f4, f10 + * fmul8x16 f8, f10, f10 + * fand f10, f6, f10 + * for f0, f2, f12 + * fpsub16 f12, f10, f10 */ #define ATTR_ALIGN(alignd) __attribute__ ((aligned(alignd))) @@ -68,1979 +68,1979 @@ static const int8_t constants_7f[] ATTR_ALIGN(8) = DUP8 (0x7f); static const int8_t constants128[] ATTR_ALIGN(8) = DUP8 (128); static const int16_t constants256_512[] ATTR_ALIGN(8) = - {256, 512, 256, 512}; + {256, 512, 256, 512}; static const int16_t constants256_1024[] ATTR_ALIGN(8) = - {256, 1024, 256, 1024}; - -#define REF_0 0 -#define REF_0_1 1 -#define REF_2 2 -#define REF_2_1 3 -#define REF_4 4 -#define REF_4_1 5 -#define REF_6 6 -#define REF_6_1 7 -#define REF_S0 8 -#define REF_S0_1 9 -#define REF_S2 10 -#define REF_S2_1 11 -#define REF_S4 12 -#define REF_S4_1 13 -#define REF_S6 14 -#define REF_S6_1 15 -#define DST_0 16 -#define DST_1 17 -#define DST_2 18 -#define DST_3 19 -#define CONST_1 20 -#define CONST_2 20 -#define CONST_3 20 -#define CONST_6 20 -#define MASK_fe 20 -#define CONST_128 22 -#define CONST_256 22 -#define CONST_512 22 -#define CONST_1024 22 -#define TMP0 24 -#define TMP1 25 -#define TMP2 26 -#define TMP3 27 -#define TMP4 28 -#define TMP5 29 -#define ZERO 30 -#define MASK_7f 30 - -#define TMP6 32 -#define TMP8 34 -#define TMP10 36 -#define TMP12 38 -#define TMP14 40 -#define TMP16 42 -#define TMP18 44 -#define TMP20 46 -#define TMP22 48 -#define TMP24 50 -#define TMP26 52 -#define TMP28 54 -#define TMP30 56 -#define TMP32 58 + {256, 1024, 256, 1024}; + +#define REF_0 0 +#define REF_0_1 1 +#define REF_2 2 +#define REF_2_1 3 +#define REF_4 4 +#define REF_4_1 5 +#define REF_6 6 +#define REF_6_1 7 +#define REF_S0 8 +#define REF_S0_1 9 +#define REF_S2 10 +#define REF_S2_1 11 +#define REF_S4 12 +#define REF_S4_1 13 +#define REF_S6 14 +#define REF_S6_1 15 +#define DST_0 16 +#define DST_1 17 +#define DST_2 18 +#define DST_3 19 +#define CONST_1 20 +#define CONST_2 20 +#define CONST_3 20 +#define CONST_6 20 +#define MASK_fe 20 +#define CONST_128 22 +#define CONST_256 22 +#define CONST_512 22 +#define CONST_1024 22 +#define TMP0 24 +#define TMP1 25 +#define TMP2 26 +#define TMP3 27 +#define TMP4 28 +#define TMP5 29 +#define ZERO 30 +#define MASK_7f 30 + +#define TMP6 32 +#define TMP8 34 +#define TMP10 36 +#define TMP12 38 +#define TMP14 40 +#define TMP16 42 +#define TMP18 44 +#define TMP20 46 +#define TMP22 48 +#define TMP24 50 +#define TMP26 52 +#define TMP28 54 +#define TMP30 56 +#define TMP32 58 static void MC_put_o_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - do { /* 5 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - - vis_faligndata(TMP0, TMP2, REF_0); - vis_st64(REF_0, dest[0]); - - vis_faligndata(TMP2, TMP4, REF_2); - vis_st64_2(REF_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + do { /* 5 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + + vis_faligndata(TMP0, TMP2, REF_0); + vis_st64(REF_0, dest[0]); + + vis_faligndata(TMP2, TMP4, REF_2); + vis_st64_2(REF_2, dest, 8); + dest += stride; + } while (--height); } static void MC_put_o_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - do { /* 4 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - ref += stride; - - /* stall */ - - vis_faligndata(TMP0, TMP2, REF_0); - vis_st64(REF_0, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + do { /* 4 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + ref += stride; + + /* stall */ + + vis_faligndata(TMP0, TMP2, REF_0); + vis_st64(REF_0, dest[0]); + dest += stride; + } while (--height); } static void MC_avg_o_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(dest[0], DST_0); - - vis_ld64(dest[8], DST_2); - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP2, TMP4, REF_2); - - vis_ld64(constants128[0], CONST_128); - - ref += stride; - height = (height >> 1) - 1; - - do { /* 24 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP6, MASK_fe, TMP6); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_xor(DST_2, REF_2, TMP8); - - vis_and(TMP8, MASK_fe, TMP8); - - vis_or(DST_0, REF_0, TMP10); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP8, TMP8); - - vis_or(DST_2, REF_2, TMP12); - vis_ld64_2(dest, stride_8, DST_2); - - vis_ld64(ref[0], TMP14); - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - - dest += stride; - vis_ld64_2(ref, 8, TMP16); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 16, TMP18); - vis_faligndata(TMP2, TMP4, REF_2); - ref += stride; - - vis_xor(DST_0, REF_0, TMP20); - - vis_and(TMP20, MASK_fe, TMP20); - - vis_xor(DST_2, REF_2, TMP22); - vis_mul8x16(CONST_128, TMP20, TMP20); - - vis_and(TMP22, MASK_fe, TMP22); - - vis_or(DST_0, REF_0, TMP24); - vis_mul8x16(CONST_128, TMP22, TMP22); - - vis_or(DST_2, REF_2, TMP26); - - vis_ld64_2(dest, stride, DST_0); - vis_faligndata(TMP14, TMP16, REF_0); - - vis_ld64_2(dest, stride_8, DST_2); - vis_faligndata(TMP16, TMP18, REF_2); - - vis_and(TMP20, MASK_7f, TMP20); - - vis_and(TMP22, MASK_7f, TMP22); - - vis_psub16(TMP24, TMP20, TMP20); - vis_st64(TMP20, dest[0]); - - vis_psub16(TMP26, TMP22, TMP22); - vis_st64_2(TMP22, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP6, MASK_fe, TMP6); - - vis_ld64_2(ref, 16, TMP4); - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_xor(DST_2, REF_2, TMP8); - - vis_and(TMP8, MASK_fe, TMP8); - - vis_or(DST_0, REF_0, TMP10); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP8, TMP8); - - vis_or(DST_2, REF_2, TMP12); - vis_ld64_2(dest, stride_8, DST_2); - - vis_ld64(ref[0], TMP14); - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - - dest += stride; - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_2); - - vis_xor(DST_0, REF_0, TMP20); - - vis_and(TMP20, MASK_fe, TMP20); - - vis_xor(DST_2, REF_2, TMP22); - vis_mul8x16(CONST_128, TMP20, TMP20); - - vis_and(TMP22, MASK_fe, TMP22); - - vis_or(DST_0, REF_0, TMP24); - vis_mul8x16(CONST_128, TMP22, TMP22); - - vis_or(DST_2, REF_2, TMP26); - - vis_and(TMP20, MASK_7f, TMP20); - - vis_and(TMP22, MASK_7f, TMP22); - - vis_psub16(TMP24, TMP20, TMP20); - vis_st64(TMP20, dest[0]); - - vis_psub16(TMP26, TMP22, TMP22); - vis_st64_2(TMP22, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(dest[0], DST_0); + + vis_ld64(dest[8], DST_2); + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP2, TMP4, REF_2); + + vis_ld64(constants128[0], CONST_128); + + ref += stride; + height = (height >> 1) - 1; + + do { /* 24 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP6, MASK_fe, TMP6); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_xor(DST_2, REF_2, TMP8); + + vis_and(TMP8, MASK_fe, TMP8); + + vis_or(DST_0, REF_0, TMP10); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP8, TMP8); + + vis_or(DST_2, REF_2, TMP12); + vis_ld64_2(dest, stride_8, DST_2); + + vis_ld64(ref[0], TMP14); + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + + dest += stride; + vis_ld64_2(ref, 8, TMP16); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 16, TMP18); + vis_faligndata(TMP2, TMP4, REF_2); + ref += stride; + + vis_xor(DST_0, REF_0, TMP20); + + vis_and(TMP20, MASK_fe, TMP20); + + vis_xor(DST_2, REF_2, TMP22); + vis_mul8x16(CONST_128, TMP20, TMP20); + + vis_and(TMP22, MASK_fe, TMP22); + + vis_or(DST_0, REF_0, TMP24); + vis_mul8x16(CONST_128, TMP22, TMP22); + + vis_or(DST_2, REF_2, TMP26); + + vis_ld64_2(dest, stride, DST_0); + vis_faligndata(TMP14, TMP16, REF_0); + + vis_ld64_2(dest, stride_8, DST_2); + vis_faligndata(TMP16, TMP18, REF_2); + + vis_and(TMP20, MASK_7f, TMP20); + + vis_and(TMP22, MASK_7f, TMP22); + + vis_psub16(TMP24, TMP20, TMP20); + vis_st64(TMP20, dest[0]); + + vis_psub16(TMP26, TMP22, TMP22); + vis_st64_2(TMP22, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP6, MASK_fe, TMP6); + + vis_ld64_2(ref, 16, TMP4); + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_xor(DST_2, REF_2, TMP8); + + vis_and(TMP8, MASK_fe, TMP8); + + vis_or(DST_0, REF_0, TMP10); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP8, TMP8); + + vis_or(DST_2, REF_2, TMP12); + vis_ld64_2(dest, stride_8, DST_2); + + vis_ld64(ref[0], TMP14); + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + + dest += stride; + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_2); + + vis_xor(DST_0, REF_0, TMP20); + + vis_and(TMP20, MASK_fe, TMP20); + + vis_xor(DST_2, REF_2, TMP22); + vis_mul8x16(CONST_128, TMP20, TMP20); + + vis_and(TMP22, MASK_fe, TMP22); + + vis_or(DST_0, REF_0, TMP24); + vis_mul8x16(CONST_128, TMP22, TMP22); + + vis_or(DST_2, REF_2, TMP26); + + vis_and(TMP20, MASK_7f, TMP20); + + vis_and(TMP22, MASK_7f, TMP22); + + vis_psub16(TMP24, TMP20, TMP20); + vis_st64(TMP20, dest[0]); + + vis_psub16(TMP26, TMP22, TMP22); + vis_st64_2(TMP22, dest, 8); } static void MC_avg_o_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(dest[0], DST_0); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants128[0], CONST_128); - - ref += stride; - height = (height >> 1) - 1; - - do { /* 12 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP4); - - vis_ld64(ref[8], TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_or(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - ref += stride; - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_ld64(ref[0], TMP12); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(ref[8], TMP2); - vis_xor(DST_0, REF_0, TMP0); - ref += stride; - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_psub16(TMP6, TMP4, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_or(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - - vis_faligndata(TMP12, TMP2, REF_0); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_psub16(TMP6, TMP0, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP4); - - vis_ld64(ref[8], TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_or(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_xor(DST_0, REF_0, TMP0); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_psub16(TMP6, TMP4, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_or(DST_0, REF_0, TMP6); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_psub16(TMP6, TMP0, TMP4); - vis_st64(TMP4, dest[0]); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(dest[0], DST_0); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants128[0], CONST_128); + + ref += stride; + height = (height >> 1) - 1; + + do { /* 12 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP4); + + vis_ld64(ref[8], TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_or(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + ref += stride; + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_ld64(ref[0], TMP12); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(ref[8], TMP2); + vis_xor(DST_0, REF_0, TMP0); + ref += stride; + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_psub16(TMP6, TMP4, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_or(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + + vis_faligndata(TMP12, TMP2, REF_0); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_psub16(TMP6, TMP0, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP4); + + vis_ld64(ref[8], TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_or(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_xor(DST_0, REF_0, TMP0); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_psub16(TMP6, TMP4, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_or(DST_0, REF_0, TMP6); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_psub16(TMP6, TMP0, TMP4); + vis_st64(TMP4, dest[0]); } static void MC_put_x_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants128[0], CONST_128); - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - ref += stride; - height = (height >> 1) - 1; - - do { /* 34 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP8); - - vis_ld64_2(ref, 16, TMP4); - vis_and(TMP6, MASK_fe, TMP6); - ref += stride; - - vis_ld64(ref[0], TMP14); - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_ld64_2(ref, 8, TMP16); - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_or(REF_0, REF_2, TMP10); - - vis_ld64_2(ref, 16, TMP18); - ref += stride; - vis_or(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - - vis_xor(REF_0, REF_2, TMP6); - - vis_xor(REF_4, REF_6, TMP8); - - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_or(REF_0, REF_2, TMP10); - - vis_or(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP14, TMP16, REF_0); - - vis_faligndata(TMP16, TMP18, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP14, TMP16, REF_2); - vis_faligndata(TMP16, TMP18, REF_6); - } else { - vis_src1(TMP16, REF_2); - vis_src1(TMP18, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP8); - - vis_ld64_2(ref, 16, TMP4); - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_or(REF_0, REF_2, TMP10); - - vis_or(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - - vis_xor(REF_0, REF_2, TMP6); - - vis_xor(REF_4, REF_6, TMP8); - - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_or(REF_0, REF_2, TMP10); - - vis_or(REF_4, REF_6, TMP12); - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_psub16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_psub16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants128[0], CONST_128); + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + ref += stride; + height = (height >> 1) - 1; + + do { /* 34 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP8); + + vis_ld64_2(ref, 16, TMP4); + vis_and(TMP6, MASK_fe, TMP6); + ref += stride; + + vis_ld64(ref[0], TMP14); + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_ld64_2(ref, 8, TMP16); + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_or(REF_0, REF_2, TMP10); + + vis_ld64_2(ref, 16, TMP18); + ref += stride; + vis_or(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + + vis_xor(REF_0, REF_2, TMP6); + + vis_xor(REF_4, REF_6, TMP8); + + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_or(REF_0, REF_2, TMP10); + + vis_or(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP14, TMP16, REF_0); + + vis_faligndata(TMP16, TMP18, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP14, TMP16, REF_2); + vis_faligndata(TMP16, TMP18, REF_6); + } else { + vis_src1(TMP16, REF_2); + vis_src1(TMP18, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP8); + + vis_ld64_2(ref, 16, TMP4); + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_or(REF_0, REF_2, TMP10); + + vis_or(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + + vis_xor(REF_0, REF_2, TMP6); + + vis_xor(REF_4, REF_6, TMP8); + + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_or(REF_0, REF_2, TMP10); + + vis_or(REF_4, REF_6, TMP12); + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_psub16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_psub16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); } static void MC_put_x_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - - vis_ld64(constants128[0], CONST_128); - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - ref += stride; - height = (height >> 1) - 1; - - do { /* 20 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - ref += stride; - - vis_ld64(ref[0], TMP8); - vis_or(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, 8, TMP10); - ref += stride; - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - vis_and(TMP4, MASK_7f, TMP4); - - vis_psub16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_or(REF_0, REF_2, TMP14); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_alignaddr_g0((void *)off); - vis_faligndata(TMP8, TMP10, REF_0); - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP8, TMP10, REF_2); - } else { - vis_src1(TMP10, REF_2); - } - - vis_and(TMP12, MASK_7f, TMP12); - - vis_psub16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_or(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - vis_and(TMP4, MASK_7f, TMP4); - - vis_psub16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_or(REF_0, REF_2, TMP14); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_psub16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + + vis_ld64(constants128[0], CONST_128); + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + ref += stride; + height = (height >> 1) - 1; + + do { /* 20 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + ref += stride; + + vis_ld64(ref[0], TMP8); + vis_or(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, 8, TMP10); + ref += stride; + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + vis_and(TMP4, MASK_7f, TMP4); + + vis_psub16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_or(REF_0, REF_2, TMP14); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_alignaddr_g0((void *)off); + vis_faligndata(TMP8, TMP10, REF_0); + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP8, TMP10, REF_2); + } else { + vis_src1(TMP10, REF_2); + } + + vis_and(TMP12, MASK_7f, TMP12); + + vis_psub16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_or(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + vis_and(TMP4, MASK_7f, TMP4); + + vis_psub16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_or(REF_0, REF_2, TMP14); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_psub16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; } static void MC_avg_x_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - vis_ld64(constants3[0], CONST_3); - vis_fzero(ZERO); - vis_ld64(constants256_512[0], CONST_256); - - ref = vis_alignaddr(ref); - do { /* 26 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_alignaddr_g0((void *)off); - - vis_ld64(ref[16], TMP4); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(dest[8], DST_2); - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_mul8x16au(REF_0, CONST_256, TMP0); - - vis_pmerge(ZERO, REF_2, TMP4); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_mul8x16al(DST_0, CONST_512, TMP4); - vis_padd16(TMP2, TMP6, TMP2); - - vis_mul8x16al(DST_1, CONST_512, TMP6); - - vis_mul8x16au(REF_6, CONST_256, TMP12); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_6_1, CONST_256, TMP14); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4, CONST_256, TMP16); - - vis_padd16(TMP0, CONST_3, TMP8); - vis_mul8x16au(REF_4_1, CONST_256, TMP18); - - vis_padd16(TMP2, CONST_3, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_padd16(TMP16, TMP12, TMP0); - - vis_st64(DST_0, dest[0]); - vis_mul8x16al(DST_2, CONST_512, TMP4); - vis_padd16(TMP18, TMP14, TMP2); - - vis_mul8x16al(DST_3, CONST_512, TMP6); - vis_padd16(TMP0, CONST_3, TMP0); - - vis_padd16(TMP2, CONST_3, TMP2); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_padd16(TMP2, TMP6, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64(DST_2, dest[8]); - - ref += stride; - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + vis_ld64(constants3[0], CONST_3); + vis_fzero(ZERO); + vis_ld64(constants256_512[0], CONST_256); + + ref = vis_alignaddr(ref); + do { /* 26 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_alignaddr_g0((void *)off); + + vis_ld64(ref[16], TMP4); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(dest[8], DST_2); + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_mul8x16au(REF_0, CONST_256, TMP0); + + vis_pmerge(ZERO, REF_2, TMP4); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_mul8x16al(DST_0, CONST_512, TMP4); + vis_padd16(TMP2, TMP6, TMP2); + + vis_mul8x16al(DST_1, CONST_512, TMP6); + + vis_mul8x16au(REF_6, CONST_256, TMP12); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_6_1, CONST_256, TMP14); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4, CONST_256, TMP16); + + vis_padd16(TMP0, CONST_3, TMP8); + vis_mul8x16au(REF_4_1, CONST_256, TMP18); + + vis_padd16(TMP2, CONST_3, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_padd16(TMP16, TMP12, TMP0); + + vis_st64(DST_0, dest[0]); + vis_mul8x16al(DST_2, CONST_512, TMP4); + vis_padd16(TMP18, TMP14, TMP2); + + vis_mul8x16al(DST_3, CONST_512, TMP6); + vis_padd16(TMP0, CONST_3, TMP0); + + vis_padd16(TMP2, CONST_3, TMP2); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_padd16(TMP2, TMP6, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64(DST_2, dest[8]); + + ref += stride; + dest += stride; + } while (--height); } static void MC_avg_x_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_times_2 = stride << 1; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - vis_ld64(constants3[0], CONST_3); - vis_fzero(ZERO); - vis_ld64(constants256_512[0], CONST_256); - - ref = vis_alignaddr(ref); - height >>= 2; - do { /* 47 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - - vis_alignaddr_g0((void *)off); - - vis_ld64(ref[0], TMP4); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 8, TMP6); - ref += stride; - - vis_ld64(ref[0], TMP8); - - vis_ld64_2(ref, 8, TMP10); - ref += stride; - vis_faligndata(TMP4, TMP6, REF_4); - - vis_ld64(ref[0], TMP12); - - vis_ld64_2(ref, 8, TMP14); - ref += stride; - vis_faligndata(TMP8, TMP10, REF_S0); - - vis_faligndata(TMP12, TMP14, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64_2(dest, stride, DST_2); - vis_faligndata(TMP4, TMP6, REF_6); - - vis_faligndata(TMP8, TMP10, REF_S2); - - vis_faligndata(TMP12, TMP14, REF_S6); - } else { - vis_ld64(dest[0], DST_0); - vis_src1(TMP2, REF_2); - - vis_ld64_2(dest, stride, DST_2); - vis_src1(TMP6, REF_6); - - vis_src1(TMP10, REF_S2); - - vis_src1(TMP14, REF_S6); - } - - vis_pmerge(ZERO, REF_0, TMP0); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_pmerge(ZERO, REF_2, TMP4); - vis_mul8x16au(REF_2_1, CONST_256, TMP6); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16al(DST_0, CONST_512, TMP16); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16al(DST_1, CONST_512, TMP18); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_4, CONST_256, TMP8); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4_1, CONST_256, TMP10); - - vis_padd16(TMP0, TMP16, TMP0); - vis_mul8x16au(REF_6, CONST_256, TMP12); - - vis_padd16(TMP2, TMP18, TMP2); - vis_mul8x16au(REF_6_1, CONST_256, TMP14); - - vis_padd16(TMP8, CONST_3, TMP8); - vis_mul8x16al(DST_2, CONST_512, TMP16); - - vis_padd16(TMP8, TMP12, TMP8); - vis_mul8x16al(DST_3, CONST_512, TMP18); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP0, DST_0); - - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP10, CONST_3, TMP10); - - vis_ld64_2(dest, stride, DST_0); - vis_padd16(TMP8, TMP16, TMP8); - - vis_ld64_2(dest, stride_times_2, TMP4/*DST_2*/); - vis_padd16(TMP10, TMP18, TMP10); - vis_pack16(TMP8, DST_2); - - vis_pack16(TMP10, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - - vis_mul8x16au(REF_S0_1, CONST_256, TMP2); - vis_pmerge(ZERO, REF_S0, TMP0); - - vis_pmerge(ZERO, REF_S2, TMP24); - vis_mul8x16au(REF_S2_1, CONST_256, TMP6); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16au(REF_S4, CONST_256, TMP8); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16au(REF_S4_1, CONST_256, TMP10); - - vis_padd16(TMP0, TMP24, TMP0); - vis_mul8x16au(REF_S6, CONST_256, TMP12); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_S6_1, CONST_256, TMP14); - - vis_padd16(TMP8, CONST_3, TMP8); - vis_mul8x16al(DST_0, CONST_512, TMP16); - - vis_padd16(TMP10, CONST_3, TMP10); - vis_mul8x16al(DST_1, CONST_512, TMP18); - - vis_padd16(TMP8, TMP12, TMP8); - vis_mul8x16al(TMP4/*DST_2*/, CONST_512, TMP20); - - vis_mul8x16al(TMP5/*DST_3*/, CONST_512, TMP22); - vis_padd16(TMP0, TMP16, TMP0); - - vis_padd16(TMP2, TMP18, TMP2); - vis_pack16(TMP0, DST_0); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_padd16(TMP8, TMP20, TMP8); - - vis_padd16(TMP10, TMP22, TMP10); - vis_pack16(TMP8, DST_2); - - vis_pack16(TMP10, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_times_2 = stride << 1; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + vis_ld64(constants3[0], CONST_3); + vis_fzero(ZERO); + vis_ld64(constants256_512[0], CONST_256); + + ref = vis_alignaddr(ref); + height >>= 2; + do { /* 47 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + + vis_alignaddr_g0((void *)off); + + vis_ld64(ref[0], TMP4); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 8, TMP6); + ref += stride; + + vis_ld64(ref[0], TMP8); + + vis_ld64_2(ref, 8, TMP10); + ref += stride; + vis_faligndata(TMP4, TMP6, REF_4); + + vis_ld64(ref[0], TMP12); + + vis_ld64_2(ref, 8, TMP14); + ref += stride; + vis_faligndata(TMP8, TMP10, REF_S0); + + vis_faligndata(TMP12, TMP14, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64_2(dest, stride, DST_2); + vis_faligndata(TMP4, TMP6, REF_6); + + vis_faligndata(TMP8, TMP10, REF_S2); + + vis_faligndata(TMP12, TMP14, REF_S6); + } else { + vis_ld64(dest[0], DST_0); + vis_src1(TMP2, REF_2); + + vis_ld64_2(dest, stride, DST_2); + vis_src1(TMP6, REF_6); + + vis_src1(TMP10, REF_S2); + + vis_src1(TMP14, REF_S6); + } + + vis_pmerge(ZERO, REF_0, TMP0); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_pmerge(ZERO, REF_2, TMP4); + vis_mul8x16au(REF_2_1, CONST_256, TMP6); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16al(DST_0, CONST_512, TMP16); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16al(DST_1, CONST_512, TMP18); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_4, CONST_256, TMP8); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4_1, CONST_256, TMP10); + + vis_padd16(TMP0, TMP16, TMP0); + vis_mul8x16au(REF_6, CONST_256, TMP12); + + vis_padd16(TMP2, TMP18, TMP2); + vis_mul8x16au(REF_6_1, CONST_256, TMP14); + + vis_padd16(TMP8, CONST_3, TMP8); + vis_mul8x16al(DST_2, CONST_512, TMP16); + + vis_padd16(TMP8, TMP12, TMP8); + vis_mul8x16al(DST_3, CONST_512, TMP18); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP0, DST_0); + + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP10, CONST_3, TMP10); + + vis_ld64_2(dest, stride, DST_0); + vis_padd16(TMP8, TMP16, TMP8); + + vis_ld64_2(dest, stride_times_2, TMP4/*DST_2*/); + vis_padd16(TMP10, TMP18, TMP10); + vis_pack16(TMP8, DST_2); + + vis_pack16(TMP10, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + + vis_mul8x16au(REF_S0_1, CONST_256, TMP2); + vis_pmerge(ZERO, REF_S0, TMP0); + + vis_pmerge(ZERO, REF_S2, TMP24); + vis_mul8x16au(REF_S2_1, CONST_256, TMP6); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16au(REF_S4, CONST_256, TMP8); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16au(REF_S4_1, CONST_256, TMP10); + + vis_padd16(TMP0, TMP24, TMP0); + vis_mul8x16au(REF_S6, CONST_256, TMP12); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_S6_1, CONST_256, TMP14); + + vis_padd16(TMP8, CONST_3, TMP8); + vis_mul8x16al(DST_0, CONST_512, TMP16); + + vis_padd16(TMP10, CONST_3, TMP10); + vis_mul8x16al(DST_1, CONST_512, TMP18); + + vis_padd16(TMP8, TMP12, TMP8); + vis_mul8x16al(TMP4/*DST_2*/, CONST_512, TMP20); + + vis_mul8x16al(TMP5/*DST_3*/, CONST_512, TMP22); + vis_padd16(TMP0, TMP16, TMP0); + + vis_padd16(TMP2, TMP18, TMP2); + vis_pack16(TMP0, DST_0); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_padd16(TMP8, TMP20, TMP8); + + vis_padd16(TMP10, TMP22, TMP10); + vis_pack16(TMP8, DST_2); + + vis_pack16(TMP10, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_put_y_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - - vis_ld64(ref[0], TMP6); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 8, TMP8); - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64_2(ref, 16, TMP10); - ref += stride; - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP6, TMP8, REF_2); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP8, TMP10, REF_6); - - vis_ld64(constants128[0], CONST_128); - height = (height >> 1) - 1; - do { /* 24 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP12); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP16); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - vis_or(REF_0, REF_2, TMP14); - - vis_ld64(ref[0], TMP6); - vis_or(REF_4, REF_6, TMP18); - - vis_ld64_2(ref, 8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(TMP16, MASK_fe, TMP16); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_mul8x16(CONST_128, TMP16, TMP16); - vis_xor(REF_0, REF_2, TMP0); - - vis_xor(REF_4, REF_6, TMP2); - - vis_or(REF_0, REF_2, TMP20); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_and(TMP16, MASK_7f, TMP16); - - vis_psub16(TMP14, TMP12, TMP12); - vis_st64(TMP12, dest[0]); - - vis_psub16(TMP18, TMP16, TMP16); - vis_st64_2(TMP16, dest, 8); - dest += stride; - - vis_or(REF_4, REF_6, TMP18); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP2, MASK_fe, TMP2); - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_faligndata(TMP6, TMP8, REF_2); - vis_mul8x16(CONST_128, TMP2, TMP2); - - vis_faligndata(TMP8, TMP10, REF_6); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_and(TMP2, MASK_7f, TMP2); - - vis_psub16(TMP20, TMP0, TMP0); - vis_st64(TMP0, dest[0]); - - vis_psub16(TMP18, TMP2, TMP2); - vis_st64_2(TMP2, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP12); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP16); - - vis_ld64_2(ref, 16, TMP4); - vis_or(REF_0, REF_2, TMP14); - - vis_or(REF_4, REF_6, TMP18); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(TMP16, MASK_fe, TMP16); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_mul8x16(CONST_128, TMP16, TMP16); - vis_xor(REF_0, REF_2, TMP0); - - vis_xor(REF_4, REF_6, TMP2); - - vis_or(REF_0, REF_2, TMP20); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_and(TMP16, MASK_7f, TMP16); - - vis_psub16(TMP14, TMP12, TMP12); - vis_st64(TMP12, dest[0]); - - vis_psub16(TMP18, TMP16, TMP16); - vis_st64_2(TMP16, dest, 8); - dest += stride; - - vis_or(REF_4, REF_6, TMP18); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP2, MASK_fe, TMP2); - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_mul8x16(CONST_128, TMP2, TMP2); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_and(TMP2, MASK_7f, TMP2); - - vis_psub16(TMP20, TMP0, TMP0); - vis_st64(TMP0, dest[0]); - - vis_psub16(TMP18, TMP2, TMP2); - vis_st64_2(TMP2, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + + vis_ld64(ref[0], TMP6); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 8, TMP8); + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64_2(ref, 16, TMP10); + ref += stride; + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP6, TMP8, REF_2); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP8, TMP10, REF_6); + + vis_ld64(constants128[0], CONST_128); + height = (height >> 1) - 1; + do { /* 24 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP12); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP16); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + vis_or(REF_0, REF_2, TMP14); + + vis_ld64(ref[0], TMP6); + vis_or(REF_4, REF_6, TMP18); + + vis_ld64_2(ref, 8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(TMP16, MASK_fe, TMP16); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_mul8x16(CONST_128, TMP16, TMP16); + vis_xor(REF_0, REF_2, TMP0); + + vis_xor(REF_4, REF_6, TMP2); + + vis_or(REF_0, REF_2, TMP20); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_and(TMP16, MASK_7f, TMP16); + + vis_psub16(TMP14, TMP12, TMP12); + vis_st64(TMP12, dest[0]); + + vis_psub16(TMP18, TMP16, TMP16); + vis_st64_2(TMP16, dest, 8); + dest += stride; + + vis_or(REF_4, REF_6, TMP18); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP2, MASK_fe, TMP2); + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_faligndata(TMP6, TMP8, REF_2); + vis_mul8x16(CONST_128, TMP2, TMP2); + + vis_faligndata(TMP8, TMP10, REF_6); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_and(TMP2, MASK_7f, TMP2); + + vis_psub16(TMP20, TMP0, TMP0); + vis_st64(TMP0, dest[0]); + + vis_psub16(TMP18, TMP2, TMP2); + vis_st64_2(TMP2, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP12); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP16); + + vis_ld64_2(ref, 16, TMP4); + vis_or(REF_0, REF_2, TMP14); + + vis_or(REF_4, REF_6, TMP18); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(TMP16, MASK_fe, TMP16); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_mul8x16(CONST_128, TMP16, TMP16); + vis_xor(REF_0, REF_2, TMP0); + + vis_xor(REF_4, REF_6, TMP2); + + vis_or(REF_0, REF_2, TMP20); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_and(TMP16, MASK_7f, TMP16); + + vis_psub16(TMP14, TMP12, TMP12); + vis_st64(TMP12, dest[0]); + + vis_psub16(TMP18, TMP16, TMP16); + vis_st64_2(TMP16, dest, 8); + dest += stride; + + vis_or(REF_4, REF_6, TMP18); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP2, MASK_fe, TMP2); + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_mul8x16(CONST_128, TMP2, TMP2); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_and(TMP2, MASK_7f, TMP2); + + vis_psub16(TMP20, TMP0, TMP0); + vis_st64(TMP0, dest[0]); + + vis_psub16(TMP18, TMP2, TMP2); + vis_st64_2(TMP2, dest, 8); } static void MC_put_y_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - - vis_ld64(ref[0], TMP4); - - vis_ld64_2(ref, 8, TMP6); - ref += stride; - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP4, TMP6, REF_2); - - vis_ld64(constants128[0], CONST_128); - height = (height >> 1) - 1; - do { /* 12 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - vis_and(TMP4, MASK_fe, TMP4); - - vis_or(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_mul8x16(CONST_128, TMP12, TMP12); - vis_or(REF_0, REF_2, TMP14); - - vis_psub16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_faligndata(TMP0, TMP2, REF_2); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_psub16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_or(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_mul8x16(CONST_128, TMP12, TMP12); - vis_or(REF_0, REF_2, TMP14); - - vis_psub16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_and(TMP12, MASK_7f, TMP12); - - vis_psub16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + + vis_ld64(ref[0], TMP4); + + vis_ld64_2(ref, 8, TMP6); + ref += stride; + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP4, TMP6, REF_2); + + vis_ld64(constants128[0], CONST_128); + height = (height >> 1) - 1; + do { /* 12 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + vis_and(TMP4, MASK_fe, TMP4); + + vis_or(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_mul8x16(CONST_128, TMP12, TMP12); + vis_or(REF_0, REF_2, TMP14); + + vis_psub16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_faligndata(TMP0, TMP2, REF_2); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_psub16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_or(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_mul8x16(CONST_128, TMP12, TMP12); + vis_or(REF_0, REF_2, TMP14); + + vis_psub16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_and(TMP12, MASK_7f, TMP12); + + vis_psub16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); } static void MC_avg_y_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants3[0], CONST_3); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_6); - height >>= 1; - - do { /* 31 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_pmerge(ZERO, REF_2, TMP12); - vis_mul8x16au(REF_2_1, CONST_256, TMP14); - - vis_ld64_2(ref, stride_8, TMP2); - vis_pmerge(ZERO, REF_6, TMP16); - vis_mul8x16au(REF_6_1, CONST_256, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(dest, 8, DST_2); - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64_2(ref, stride, TMP6); - vis_pmerge(ZERO, REF_0, TMP0); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_ld64_2(ref, stride_8, TMP8); - vis_pmerge(ZERO, REF_4, TMP4); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - - vis_ld64_2(dest, stride, REF_S0/*DST_4*/); - vis_faligndata(TMP6, TMP8, REF_2); - vis_mul8x16au(REF_4_1, CONST_256, TMP6); - - vis_ld64_2(dest, stride_8, REF_S2/*DST_6*/); - vis_faligndata(TMP8, TMP10, REF_6); - vis_mul8x16al(DST_0, CONST_512, TMP20); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16al(DST_1, CONST_512, TMP22); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16al(DST_2, CONST_512, TMP24); - - vis_padd16(TMP4, CONST_3, TMP4); - vis_mul8x16al(DST_3, CONST_512, TMP26); - - vis_padd16(TMP6, CONST_3, TMP6); - - vis_padd16(TMP12, TMP20, TMP12); - vis_mul8x16al(REF_S0, CONST_512, TMP20); - - vis_padd16(TMP14, TMP22, TMP14); - vis_mul8x16al(REF_S0_1, CONST_512, TMP22); - - vis_padd16(TMP16, TMP24, TMP16); - vis_mul8x16al(REF_S2, CONST_512, TMP24); - - vis_padd16(TMP18, TMP26, TMP18); - vis_mul8x16al(REF_S2_1, CONST_512, TMP26); - - vis_padd16(TMP12, TMP0, TMP12); - vis_mul8x16au(REF_2, CONST_256, TMP28); - - vis_padd16(TMP14, TMP2, TMP14); - vis_mul8x16au(REF_2_1, CONST_256, TMP30); - - vis_padd16(TMP16, TMP4, TMP16); - vis_mul8x16au(REF_6, CONST_256, REF_S4); - - vis_padd16(TMP18, TMP6, TMP18); - vis_mul8x16au(REF_6_1, CONST_256, REF_S6); - - vis_pack16(TMP12, DST_0); - vis_padd16(TMP28, TMP0, TMP12); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP30, TMP2, TMP14); - - vis_pack16(TMP16, DST_2); - vis_padd16(REF_S4, TMP4, TMP16); - - vis_pack16(TMP18, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - vis_padd16(REF_S6, TMP6, TMP18); - - vis_padd16(TMP12, TMP20, TMP12); - - vis_padd16(TMP14, TMP22, TMP14); - vis_pack16(TMP12, DST_0); - - vis_padd16(TMP16, TMP24, TMP16); - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - - vis_padd16(TMP18, TMP26, TMP18); - vis_pack16(TMP16, DST_2); - - vis_pack16(TMP18, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants3[0], CONST_3); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_6); + height >>= 1; + + do { /* 31 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_pmerge(ZERO, REF_2, TMP12); + vis_mul8x16au(REF_2_1, CONST_256, TMP14); + + vis_ld64_2(ref, stride_8, TMP2); + vis_pmerge(ZERO, REF_6, TMP16); + vis_mul8x16au(REF_6_1, CONST_256, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(dest, 8, DST_2); + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64_2(ref, stride, TMP6); + vis_pmerge(ZERO, REF_0, TMP0); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_ld64_2(ref, stride_8, TMP8); + vis_pmerge(ZERO, REF_4, TMP4); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + + vis_ld64_2(dest, stride, REF_S0/*DST_4*/); + vis_faligndata(TMP6, TMP8, REF_2); + vis_mul8x16au(REF_4_1, CONST_256, TMP6); + + vis_ld64_2(dest, stride_8, REF_S2/*DST_6*/); + vis_faligndata(TMP8, TMP10, REF_6); + vis_mul8x16al(DST_0, CONST_512, TMP20); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16al(DST_1, CONST_512, TMP22); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16al(DST_2, CONST_512, TMP24); + + vis_padd16(TMP4, CONST_3, TMP4); + vis_mul8x16al(DST_3, CONST_512, TMP26); + + vis_padd16(TMP6, CONST_3, TMP6); + + vis_padd16(TMP12, TMP20, TMP12); + vis_mul8x16al(REF_S0, CONST_512, TMP20); + + vis_padd16(TMP14, TMP22, TMP14); + vis_mul8x16al(REF_S0_1, CONST_512, TMP22); + + vis_padd16(TMP16, TMP24, TMP16); + vis_mul8x16al(REF_S2, CONST_512, TMP24); + + vis_padd16(TMP18, TMP26, TMP18); + vis_mul8x16al(REF_S2_1, CONST_512, TMP26); + + vis_padd16(TMP12, TMP0, TMP12); + vis_mul8x16au(REF_2, CONST_256, TMP28); + + vis_padd16(TMP14, TMP2, TMP14); + vis_mul8x16au(REF_2_1, CONST_256, TMP30); + + vis_padd16(TMP16, TMP4, TMP16); + vis_mul8x16au(REF_6, CONST_256, REF_S4); + + vis_padd16(TMP18, TMP6, TMP18); + vis_mul8x16au(REF_6_1, CONST_256, REF_S6); + + vis_pack16(TMP12, DST_0); + vis_padd16(TMP28, TMP0, TMP12); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP30, TMP2, TMP14); + + vis_pack16(TMP16, DST_2); + vis_padd16(REF_S4, TMP4, TMP16); + + vis_pack16(TMP18, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + vis_padd16(REF_S6, TMP6, TMP18); + + vis_padd16(TMP12, TMP20, TMP12); + + vis_padd16(TMP14, TMP22, TMP14); + vis_pack16(TMP12, DST_0); + + vis_padd16(TMP16, TMP24, TMP16); + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + + vis_padd16(TMP18, TMP26, TMP18); + vis_pack16(TMP16, DST_2); + + vis_pack16(TMP18, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_avg_y_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(constants3[0], CONST_3); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64(constants256_512[0], CONST_256); - - height >>= 1; - do { /* 20 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_pmerge(ZERO, REF_2, TMP8); - vis_mul8x16au(REF_2_1, CONST_256, TMP10); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - - vis_ld64(dest[0], DST_0); - - vis_ld64_2(dest, stride, DST_2); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride, TMP4); - vis_mul8x16al(DST_0, CONST_512, TMP16); - vis_pmerge(ZERO, REF_0, TMP12); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - vis_mul8x16al(DST_1, CONST_512, TMP18); - vis_pmerge(ZERO, REF_0_1, TMP14); - - vis_padd16(TMP12, CONST_3, TMP12); - vis_mul8x16al(DST_2, CONST_512, TMP24); - - vis_padd16(TMP14, CONST_3, TMP14); - vis_mul8x16al(DST_3, CONST_512, TMP26); - - vis_faligndata(TMP4, TMP6, REF_2); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - vis_mul8x16au(REF_2, CONST_256, TMP20); - - vis_padd16(TMP8, TMP16, TMP0); - vis_mul8x16au(REF_2_1, CONST_256, TMP22); - - vis_padd16(TMP10, TMP18, TMP2); - vis_pack16(TMP0, DST_0); - - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP12, TMP20, TMP12); - - vis_padd16(TMP14, TMP22, TMP14); - - vis_padd16(TMP12, TMP24, TMP0); - - vis_padd16(TMP14, TMP26, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(constants3[0], CONST_3); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64(constants256_512[0], CONST_256); + + height >>= 1; + do { /* 20 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_pmerge(ZERO, REF_2, TMP8); + vis_mul8x16au(REF_2_1, CONST_256, TMP10); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + + vis_ld64(dest[0], DST_0); + + vis_ld64_2(dest, stride, DST_2); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride, TMP4); + vis_mul8x16al(DST_0, CONST_512, TMP16); + vis_pmerge(ZERO, REF_0, TMP12); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + vis_mul8x16al(DST_1, CONST_512, TMP18); + vis_pmerge(ZERO, REF_0_1, TMP14); + + vis_padd16(TMP12, CONST_3, TMP12); + vis_mul8x16al(DST_2, CONST_512, TMP24); + + vis_padd16(TMP14, CONST_3, TMP14); + vis_mul8x16al(DST_3, CONST_512, TMP26); + + vis_faligndata(TMP4, TMP6, REF_2); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + vis_mul8x16au(REF_2, CONST_256, TMP20); + + vis_padd16(TMP8, TMP16, TMP0); + vis_mul8x16au(REF_2_1, CONST_256, TMP22); + + vis_padd16(TMP10, TMP18, TMP2); + vis_pack16(TMP0, DST_0); + + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP12, TMP20, TMP12); + + vis_padd16(TMP14, TMP22, TMP14); + + vis_padd16(TMP12, TMP24, TMP0); + + vis_padd16(TMP14, TMP26, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_put_xy_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants2[0], CONST_2); - vis_faligndata(TMP0, TMP2, REF_S0); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - vis_faligndata(TMP2, TMP4, REF_S6); - } else { - vis_src1(TMP2, REF_S2); - vis_src1(TMP4, REF_S6); - } - - height >>= 1; - do { - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S0_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - vis_mul8x16au(REF_S2, CONST_256, TMP16); - vis_pmerge(ZERO, REF_S2_1, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - vis_mul8x16au(REF_S4, CONST_256, TMP20); - vis_pmerge(ZERO, REF_S4_1, TMP22); - - vis_ld64_2(ref, stride, TMP6); - vis_mul8x16au(REF_S6, CONST_256, TMP24); - vis_pmerge(ZERO, REF_S6_1, TMP26); - - vis_ld64_2(ref, stride_8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_faligndata(TMP6, TMP8, REF_S0); - - vis_faligndata(TMP8, TMP10, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - vis_faligndata(TMP6, TMP8, REF_S2); - vis_faligndata(TMP8, TMP10, REF_S6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - vis_src1(TMP8, REF_S2); - vis_src1(TMP10, REF_S6); - } - - vis_mul8x16au(REF_0, CONST_256, TMP0); - vis_pmerge(ZERO, REF_0_1, TMP2); - - vis_mul8x16au(REF_2, CONST_256, TMP4); - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_padd16(TMP0, CONST_2, TMP8); - vis_mul8x16au(REF_4, CONST_256, TMP0); - - vis_padd16(TMP2, CONST_2, TMP10); - vis_mul8x16au(REF_4_1, CONST_256, TMP2); - - vis_padd16(TMP8, TMP4, TMP8); - vis_mul8x16au(REF_6, CONST_256, TMP4); - - vis_padd16(TMP10, TMP6, TMP10); - vis_mul8x16au(REF_6_1, CONST_256, TMP6); - - vis_padd16(TMP12, TMP8, TMP12); - - vis_padd16(TMP14, TMP10, TMP14); - - vis_padd16(TMP12, TMP16, TMP12); - - vis_padd16(TMP14, TMP18, TMP14); - vis_pack16(TMP12, DST_0); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP0, CONST_2, TMP12); - - vis_mul8x16au(REF_S0, CONST_256, TMP0); - vis_padd16(TMP2, CONST_2, TMP14); - - vis_mul8x16au(REF_S0_1, CONST_256, TMP2); - vis_padd16(TMP12, TMP4, TMP12); - - vis_mul8x16au(REF_S2, CONST_256, TMP4); - vis_padd16(TMP14, TMP6, TMP14); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP6); - vis_padd16(TMP20, TMP12, TMP20); - - vis_padd16(TMP22, TMP14, TMP22); - - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP22, TMP26, TMP22); - vis_pack16(TMP20, DST_2); - - vis_pack16(TMP22, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - vis_padd16(TMP0, TMP4, TMP24); - - vis_mul8x16au(REF_S4, CONST_256, TMP0); - vis_padd16(TMP2, TMP6, TMP26); - - vis_mul8x16au(REF_S4_1, CONST_256, TMP2); - vis_padd16(TMP24, TMP8, TMP24); - - vis_padd16(TMP26, TMP10, TMP26); - vis_pack16(TMP24, DST_0); - - vis_pack16(TMP26, DST_1); - vis_st64(DST_0, dest[0]); - vis_pmerge(ZERO, REF_S6, TMP4); - - vis_pmerge(ZERO, REF_S6_1, TMP6); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_padd16(TMP2, TMP6, TMP2); - - vis_padd16(TMP0, TMP12, TMP0); - - vis_padd16(TMP2, TMP14, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants2[0], CONST_2); + vis_faligndata(TMP0, TMP2, REF_S0); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + vis_faligndata(TMP2, TMP4, REF_S6); + } else { + vis_src1(TMP2, REF_S2); + vis_src1(TMP4, REF_S6); + } + + height >>= 1; + do { + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S0_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + vis_mul8x16au(REF_S2, CONST_256, TMP16); + vis_pmerge(ZERO, REF_S2_1, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + vis_mul8x16au(REF_S4, CONST_256, TMP20); + vis_pmerge(ZERO, REF_S4_1, TMP22); + + vis_ld64_2(ref, stride, TMP6); + vis_mul8x16au(REF_S6, CONST_256, TMP24); + vis_pmerge(ZERO, REF_S6_1, TMP26); + + vis_ld64_2(ref, stride_8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_faligndata(TMP6, TMP8, REF_S0); + + vis_faligndata(TMP8, TMP10, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + vis_faligndata(TMP6, TMP8, REF_S2); + vis_faligndata(TMP8, TMP10, REF_S6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + vis_src1(TMP8, REF_S2); + vis_src1(TMP10, REF_S6); + } + + vis_mul8x16au(REF_0, CONST_256, TMP0); + vis_pmerge(ZERO, REF_0_1, TMP2); + + vis_mul8x16au(REF_2, CONST_256, TMP4); + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_padd16(TMP0, CONST_2, TMP8); + vis_mul8x16au(REF_4, CONST_256, TMP0); + + vis_padd16(TMP2, CONST_2, TMP10); + vis_mul8x16au(REF_4_1, CONST_256, TMP2); + + vis_padd16(TMP8, TMP4, TMP8); + vis_mul8x16au(REF_6, CONST_256, TMP4); + + vis_padd16(TMP10, TMP6, TMP10); + vis_mul8x16au(REF_6_1, CONST_256, TMP6); + + vis_padd16(TMP12, TMP8, TMP12); + + vis_padd16(TMP14, TMP10, TMP14); + + vis_padd16(TMP12, TMP16, TMP12); + + vis_padd16(TMP14, TMP18, TMP14); + vis_pack16(TMP12, DST_0); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP0, CONST_2, TMP12); + + vis_mul8x16au(REF_S0, CONST_256, TMP0); + vis_padd16(TMP2, CONST_2, TMP14); + + vis_mul8x16au(REF_S0_1, CONST_256, TMP2); + vis_padd16(TMP12, TMP4, TMP12); + + vis_mul8x16au(REF_S2, CONST_256, TMP4); + vis_padd16(TMP14, TMP6, TMP14); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP6); + vis_padd16(TMP20, TMP12, TMP20); + + vis_padd16(TMP22, TMP14, TMP22); + + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP22, TMP26, TMP22); + vis_pack16(TMP20, DST_2); + + vis_pack16(TMP22, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + vis_padd16(TMP0, TMP4, TMP24); + + vis_mul8x16au(REF_S4, CONST_256, TMP0); + vis_padd16(TMP2, TMP6, TMP26); + + vis_mul8x16au(REF_S4_1, CONST_256, TMP2); + vis_padd16(TMP24, TMP8, TMP24); + + vis_padd16(TMP26, TMP10, TMP26); + vis_pack16(TMP24, DST_0); + + vis_pack16(TMP26, DST_1); + vis_st64(DST_0, dest[0]); + vis_pmerge(ZERO, REF_S6, TMP4); + + vis_pmerge(ZERO, REF_S6_1, TMP6); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_padd16(TMP2, TMP6, TMP2); + + vis_padd16(TMP0, TMP12, TMP0); + + vis_padd16(TMP2, TMP14, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_put_xy_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(constants2[0], CONST_2); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP0, TMP2, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - } else { - vis_src1(TMP2, REF_S2); - } - - height >>= 1; - do { /* 26 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP8); - vis_pmerge(ZERO, REF_S2, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - vis_mul8x16au(REF_S0_1, CONST_256, TMP10); - vis_pmerge(ZERO, REF_S2_1, TMP14); - - vis_ld64_2(ref, stride, TMP4); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - vis_faligndata(TMP0, TMP2, REF_S4); - - vis_pmerge(ZERO, REF_S4, TMP18); - - vis_pmerge(ZERO, REF_S4_1, TMP20); - - vis_faligndata(TMP4, TMP6, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S6); - vis_faligndata(TMP4, TMP6, REF_S2); - } else { - vis_src1(TMP2, REF_S6); - vis_src1(TMP6, REF_S2); - } - - vis_padd16(TMP18, CONST_2, TMP18); - vis_mul8x16au(REF_S6, CONST_256, TMP22); - - vis_padd16(TMP20, CONST_2, TMP20); - vis_mul8x16au(REF_S6_1, CONST_256, TMP24); - - vis_mul8x16au(REF_S0, CONST_256, TMP26); - vis_pmerge(ZERO, REF_S0_1, TMP28); - - vis_mul8x16au(REF_S2, CONST_256, TMP30); - vis_padd16(TMP18, TMP22, TMP18); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP32); - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP8, TMP18, TMP8); - - vis_padd16(TMP10, TMP20, TMP10); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP18, TMP26, TMP18); - - vis_padd16(TMP20, TMP28, TMP20); - - vis_padd16(TMP18, TMP30, TMP18); - - vis_padd16(TMP20, TMP32, TMP20); - vis_pack16(TMP18, DST_2); - - vis_pack16(TMP20, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(constants2[0], CONST_2); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP0, TMP2, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + } else { + vis_src1(TMP2, REF_S2); + } + + height >>= 1; + do { /* 26 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP8); + vis_pmerge(ZERO, REF_S2, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + vis_mul8x16au(REF_S0_1, CONST_256, TMP10); + vis_pmerge(ZERO, REF_S2_1, TMP14); + + vis_ld64_2(ref, stride, TMP4); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + vis_faligndata(TMP0, TMP2, REF_S4); + + vis_pmerge(ZERO, REF_S4, TMP18); + + vis_pmerge(ZERO, REF_S4_1, TMP20); + + vis_faligndata(TMP4, TMP6, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S6); + vis_faligndata(TMP4, TMP6, REF_S2); + } else { + vis_src1(TMP2, REF_S6); + vis_src1(TMP6, REF_S2); + } + + vis_padd16(TMP18, CONST_2, TMP18); + vis_mul8x16au(REF_S6, CONST_256, TMP22); + + vis_padd16(TMP20, CONST_2, TMP20); + vis_mul8x16au(REF_S6_1, CONST_256, TMP24); + + vis_mul8x16au(REF_S0, CONST_256, TMP26); + vis_pmerge(ZERO, REF_S0_1, TMP28); + + vis_mul8x16au(REF_S2, CONST_256, TMP30); + vis_padd16(TMP18, TMP22, TMP18); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP32); + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP8, TMP18, TMP8); + + vis_padd16(TMP10, TMP20, TMP10); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP18, TMP26, TMP18); + + vis_padd16(TMP20, TMP28, TMP20); + + vis_padd16(TMP18, TMP30, TMP18); + + vis_padd16(TMP20, TMP32, TMP20); + vis_pack16(TMP18, DST_2); + + vis_pack16(TMP20, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_avg_xy_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants6[0], CONST_6); - vis_faligndata(TMP0, TMP2, REF_S0); - - vis_ld64(constants256_1024[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - vis_faligndata(TMP2, TMP4, REF_S6); - } else { - vis_src1(TMP2, REF_S2); - vis_src1(TMP4, REF_S6); - } - - height >>= 1; - do { /* 55 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S0_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - vis_mul8x16au(REF_S2, CONST_256, TMP16); - vis_pmerge(ZERO, REF_S2_1, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - vis_mul8x16au(REF_S4, CONST_256, TMP20); - vis_pmerge(ZERO, REF_S4_1, TMP22); - - vis_ld64_2(ref, stride, TMP6); - vis_mul8x16au(REF_S6, CONST_256, TMP24); - vis_pmerge(ZERO, REF_S6_1, TMP26); - - vis_ld64_2(ref, stride_8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP6, TMP8, REF_S0); - - vis_ld64_2(dest, 8, DST_2); - vis_faligndata(TMP8, TMP10, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - vis_faligndata(TMP6, TMP8, REF_S2); - vis_faligndata(TMP8, TMP10, REF_S6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - vis_src1(TMP8, REF_S2); - vis_src1(TMP10, REF_S6); - } - - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_0, TMP0); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_pmerge(ZERO, REF_0_1, TMP2); - - vis_mul8x16au(REF_2, CONST_256, TMP4); - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_mul8x16al(DST_2, CONST_1024, REF_0); - vis_padd16(TMP0, CONST_6, TMP0); - - vis_mul8x16al(DST_3, CONST_1024, REF_2); - vis_padd16(TMP2, CONST_6, TMP2); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_4, CONST_256, TMP4); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4_1, CONST_256, TMP6); - - vis_padd16(TMP12, TMP0, TMP12); - vis_mul8x16au(REF_6, CONST_256, TMP8); - - vis_padd16(TMP14, TMP2, TMP14); - vis_mul8x16au(REF_6_1, CONST_256, TMP10); - - vis_padd16(TMP12, TMP16, TMP12); - vis_mul8x16au(REF_S0, CONST_256, REF_4); - - vis_padd16(TMP14, TMP18, TMP14); - vis_mul8x16au(REF_S0_1, CONST_256, REF_6); - - vis_padd16(TMP12, TMP30, TMP12); - - vis_padd16(TMP14, TMP32, TMP14); - vis_pack16(TMP12, DST_0); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP4, CONST_6, TMP4); - - vis_ld64_2(dest, stride, DST_0); - vis_padd16(TMP6, CONST_6, TMP6); - vis_mul8x16au(REF_S2, CONST_256, TMP12); - - vis_padd16(TMP4, TMP8, TMP4); - vis_mul8x16au(REF_S2_1, CONST_256, TMP14); - - vis_padd16(TMP6, TMP10, TMP6); - - vis_padd16(TMP20, TMP4, TMP20); - - vis_padd16(TMP22, TMP6, TMP22); - - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP22, TMP26, TMP22); - - vis_padd16(TMP20, REF_0, TMP20); - vis_mul8x16au(REF_S4, CONST_256, REF_0); - - vis_padd16(TMP22, REF_2, TMP22); - vis_pack16(TMP20, DST_2); - - vis_pack16(TMP22, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - - vis_ld64_2(dest, 8, DST_2); - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_S4_1, REF_2); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_padd16(REF_4, TMP0, TMP8); - - vis_mul8x16au(REF_S6, CONST_256, REF_4); - vis_padd16(REF_6, TMP2, TMP10); - - vis_mul8x16au(REF_S6_1, CONST_256, REF_6); - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - - vis_padd16(TMP8, TMP30, TMP8); - - vis_padd16(TMP10, TMP32, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - - vis_padd16(REF_0, TMP4, REF_0); - - vis_mul8x16al(DST_2, CONST_1024, TMP30); - vis_padd16(REF_2, TMP6, REF_2); - - vis_mul8x16al(DST_3, CONST_1024, TMP32); - vis_padd16(REF_0, REF_4, REF_0); - - vis_padd16(REF_2, REF_6, REF_2); - - vis_padd16(REF_0, TMP30, REF_0); - - /* stall */ - - vis_padd16(REF_2, TMP32, REF_2); - vis_pack16(REF_0, DST_2); - - vis_pack16(REF_2, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants6[0], CONST_6); + vis_faligndata(TMP0, TMP2, REF_S0); + + vis_ld64(constants256_1024[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + vis_faligndata(TMP2, TMP4, REF_S6); + } else { + vis_src1(TMP2, REF_S2); + vis_src1(TMP4, REF_S6); + } + + height >>= 1; + do { /* 55 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S0_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + vis_mul8x16au(REF_S2, CONST_256, TMP16); + vis_pmerge(ZERO, REF_S2_1, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + vis_mul8x16au(REF_S4, CONST_256, TMP20); + vis_pmerge(ZERO, REF_S4_1, TMP22); + + vis_ld64_2(ref, stride, TMP6); + vis_mul8x16au(REF_S6, CONST_256, TMP24); + vis_pmerge(ZERO, REF_S6_1, TMP26); + + vis_ld64_2(ref, stride_8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP6, TMP8, REF_S0); + + vis_ld64_2(dest, 8, DST_2); + vis_faligndata(TMP8, TMP10, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + vis_faligndata(TMP6, TMP8, REF_S2); + vis_faligndata(TMP8, TMP10, REF_S6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + vis_src1(TMP8, REF_S2); + vis_src1(TMP10, REF_S6); + } + + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_0, TMP0); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_pmerge(ZERO, REF_0_1, TMP2); + + vis_mul8x16au(REF_2, CONST_256, TMP4); + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_mul8x16al(DST_2, CONST_1024, REF_0); + vis_padd16(TMP0, CONST_6, TMP0); + + vis_mul8x16al(DST_3, CONST_1024, REF_2); + vis_padd16(TMP2, CONST_6, TMP2); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_4, CONST_256, TMP4); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4_1, CONST_256, TMP6); + + vis_padd16(TMP12, TMP0, TMP12); + vis_mul8x16au(REF_6, CONST_256, TMP8); + + vis_padd16(TMP14, TMP2, TMP14); + vis_mul8x16au(REF_6_1, CONST_256, TMP10); + + vis_padd16(TMP12, TMP16, TMP12); + vis_mul8x16au(REF_S0, CONST_256, REF_4); + + vis_padd16(TMP14, TMP18, TMP14); + vis_mul8x16au(REF_S0_1, CONST_256, REF_6); + + vis_padd16(TMP12, TMP30, TMP12); + + vis_padd16(TMP14, TMP32, TMP14); + vis_pack16(TMP12, DST_0); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP4, CONST_6, TMP4); + + vis_ld64_2(dest, stride, DST_0); + vis_padd16(TMP6, CONST_6, TMP6); + vis_mul8x16au(REF_S2, CONST_256, TMP12); + + vis_padd16(TMP4, TMP8, TMP4); + vis_mul8x16au(REF_S2_1, CONST_256, TMP14); + + vis_padd16(TMP6, TMP10, TMP6); + + vis_padd16(TMP20, TMP4, TMP20); + + vis_padd16(TMP22, TMP6, TMP22); + + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP22, TMP26, TMP22); + + vis_padd16(TMP20, REF_0, TMP20); + vis_mul8x16au(REF_S4, CONST_256, REF_0); + + vis_padd16(TMP22, REF_2, TMP22); + vis_pack16(TMP20, DST_2); + + vis_pack16(TMP22, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + + vis_ld64_2(dest, 8, DST_2); + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_S4_1, REF_2); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_padd16(REF_4, TMP0, TMP8); + + vis_mul8x16au(REF_S6, CONST_256, REF_4); + vis_padd16(REF_6, TMP2, TMP10); + + vis_mul8x16au(REF_S6_1, CONST_256, REF_6); + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + + vis_padd16(TMP8, TMP30, TMP8); + + vis_padd16(TMP10, TMP32, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + + vis_padd16(REF_0, TMP4, REF_0); + + vis_mul8x16al(DST_2, CONST_1024, TMP30); + vis_padd16(REF_2, TMP6, REF_2); + + vis_mul8x16al(DST_3, CONST_1024, TMP32); + vis_padd16(REF_0, REF_4, REF_0); + + vis_padd16(REF_2, REF_6, REF_2); + + vis_padd16(REF_0, TMP30, REF_0); + + /* stall */ + + vis_padd16(REF_2, TMP32, REF_2); + vis_pack16(REF_0, DST_2); + + vis_pack16(REF_2, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_avg_xy_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - - vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - vis_fzero(ZERO); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64(constants6[0], CONST_6); - - vis_ld64(constants256_1024[0], CONST_256); - vis_faligndata(TMP0, TMP2, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - } else { - vis_src1(TMP2, REF_S2); - } - - height >>= 1; - do { /* 31 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP8); - vis_pmerge(ZERO, REF_S0_1, TMP10); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - vis_mul8x16au(REF_S2, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S2_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride, TMP4); - vis_faligndata(TMP0, TMP2, REF_S4); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP4, TMP6, REF_S0); - - vis_ld64_2(dest, stride, DST_2); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S6); - vis_faligndata(TMP4, TMP6, REF_S2); - } else { - vis_src1(TMP2, REF_S6); - vis_src1(TMP6, REF_S2); - } - - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_S4, TMP22); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_pmerge(ZERO, REF_S4_1, TMP24); - - vis_mul8x16au(REF_S6, CONST_256, TMP26); - vis_pmerge(ZERO, REF_S6_1, TMP28); - - vis_mul8x16au(REF_S0, CONST_256, REF_S4); - vis_padd16(TMP22, CONST_6, TMP22); - - vis_mul8x16au(REF_S0_1, CONST_256, REF_S6); - vis_padd16(TMP24, CONST_6, TMP24); - - vis_mul8x16al(DST_2, CONST_1024, REF_0); - vis_padd16(TMP22, TMP26, TMP22); - - vis_mul8x16al(DST_3, CONST_1024, REF_2); - vis_padd16(TMP24, TMP28, TMP24); - - vis_mul8x16au(REF_S2, CONST_256, TMP26); - vis_padd16(TMP8, TMP22, TMP8); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP28); - vis_padd16(TMP10, TMP24, TMP10); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - - vis_padd16(TMP8, TMP30, TMP8); - - vis_padd16(TMP10, TMP32, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_padd16(REF_S4, TMP22, TMP12); - - vis_padd16(REF_S6, TMP24, TMP14); - - vis_padd16(TMP12, TMP26, TMP12); - - vis_padd16(TMP14, TMP28, TMP14); - - vis_padd16(TMP12, REF_0, TMP12); - - vis_padd16(TMP14, REF_2, TMP14); - vis_pack16(TMP12, DST_2); - - vis_pack16(TMP14, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + + vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + vis_fzero(ZERO); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64(constants6[0], CONST_6); + + vis_ld64(constants256_1024[0], CONST_256); + vis_faligndata(TMP0, TMP2, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + } else { + vis_src1(TMP2, REF_S2); + } + + height >>= 1; + do { /* 31 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP8); + vis_pmerge(ZERO, REF_S0_1, TMP10); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + vis_mul8x16au(REF_S2, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S2_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride, TMP4); + vis_faligndata(TMP0, TMP2, REF_S4); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP4, TMP6, REF_S0); + + vis_ld64_2(dest, stride, DST_2); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S6); + vis_faligndata(TMP4, TMP6, REF_S2); + } else { + vis_src1(TMP2, REF_S6); + vis_src1(TMP6, REF_S2); + } + + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_S4, TMP22); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_pmerge(ZERO, REF_S4_1, TMP24); + + vis_mul8x16au(REF_S6, CONST_256, TMP26); + vis_pmerge(ZERO, REF_S6_1, TMP28); + + vis_mul8x16au(REF_S0, CONST_256, REF_S4); + vis_padd16(TMP22, CONST_6, TMP22); + + vis_mul8x16au(REF_S0_1, CONST_256, REF_S6); + vis_padd16(TMP24, CONST_6, TMP24); + + vis_mul8x16al(DST_2, CONST_1024, REF_0); + vis_padd16(TMP22, TMP26, TMP22); + + vis_mul8x16al(DST_3, CONST_1024, REF_2); + vis_padd16(TMP24, TMP28, TMP24); + + vis_mul8x16au(REF_S2, CONST_256, TMP26); + vis_padd16(TMP8, TMP22, TMP8); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP28); + vis_padd16(TMP10, TMP24, TMP10); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + + vis_padd16(TMP8, TMP30, TMP8); + + vis_padd16(TMP10, TMP32, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_padd16(REF_S4, TMP22, TMP12); + + vis_padd16(REF_S6, TMP24, TMP14); + + vis_padd16(TMP12, TMP26, TMP12); + + vis_padd16(TMP14, TMP28, TMP14); + + vis_padd16(TMP12, REF_0, TMP12); + + vis_padd16(TMP14, REF_2, TMP14); + vis_pack16(TMP12, DST_2); + + vis_pack16(TMP14, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } /* End of rounding code */ @@ -2058,1930 +2058,1930 @@ * of '0xfe' is in f4, a repeating mask of '0x7f' is in f6, and * the value 0x80808080 is in f8): * - * fxor f0, f2, f10 - * fand f10, f4, f10 - * fmul8x16 f8, f10, f10 - * fand f10, f6, f10 - * fand f0, f2, f12 - * fpadd16 f12, f10, f10 + * fxor f0, f2, f10 + * fand f10, f4, f10 + * fmul8x16 f8, f10, f10 + * fand f10, f6, f10 + * fand f0, f2, f12 + * fpadd16 f12, f10, f10 */ static void MC_put_no_round_o_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - do { /* 5 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - - vis_faligndata(TMP0, TMP2, REF_0); - vis_st64(REF_0, dest[0]); - - vis_faligndata(TMP2, TMP4, REF_2); - vis_st64_2(REF_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + do { /* 5 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + + vis_faligndata(TMP0, TMP2, REF_0); + vis_st64(REF_0, dest[0]); + + vis_faligndata(TMP2, TMP4, REF_2); + vis_st64_2(REF_2, dest, 8); + dest += stride; + } while (--height); } static void MC_put_no_round_o_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - do { /* 4 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - ref += stride; - - /* stall */ - - vis_faligndata(TMP0, TMP2, REF_0); - vis_st64(REF_0, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + do { /* 4 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + ref += stride; + + /* stall */ + + vis_faligndata(TMP0, TMP2, REF_0); + vis_st64(REF_0, dest[0]); + dest += stride; + } while (--height); } static void MC_avg_no_round_o_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(dest[0], DST_0); - - vis_ld64(dest[8], DST_2); - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP2, TMP4, REF_2); - - vis_ld64(constants128[0], CONST_128); - - ref += stride; - height = (height >> 1) - 1; - - do { /* 24 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP6, MASK_fe, TMP6); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_xor(DST_2, REF_2, TMP8); - - vis_and(TMP8, MASK_fe, TMP8); - - vis_and(DST_0, REF_0, TMP10); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP8, TMP8); - - vis_and(DST_2, REF_2, TMP12); - vis_ld64_2(dest, stride_8, DST_2); - - vis_ld64(ref[0], TMP14); - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - - dest += stride; - vis_ld64_2(ref, 8, TMP16); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 16, TMP18); - vis_faligndata(TMP2, TMP4, REF_2); - ref += stride; - - vis_xor(DST_0, REF_0, TMP20); - - vis_and(TMP20, MASK_fe, TMP20); - - vis_xor(DST_2, REF_2, TMP22); - vis_mul8x16(CONST_128, TMP20, TMP20); - - vis_and(TMP22, MASK_fe, TMP22); - - vis_and(DST_0, REF_0, TMP24); - vis_mul8x16(CONST_128, TMP22, TMP22); - - vis_and(DST_2, REF_2, TMP26); - - vis_ld64_2(dest, stride, DST_0); - vis_faligndata(TMP14, TMP16, REF_0); - - vis_ld64_2(dest, stride_8, DST_2); - vis_faligndata(TMP16, TMP18, REF_2); - - vis_and(TMP20, MASK_7f, TMP20); - - vis_and(TMP22, MASK_7f, TMP22); - - vis_padd16(TMP24, TMP20, TMP20); - vis_st64(TMP20, dest[0]); - - vis_padd16(TMP26, TMP22, TMP22); - vis_st64_2(TMP22, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP6, MASK_fe, TMP6); - - vis_ld64_2(ref, 16, TMP4); - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_xor(DST_2, REF_2, TMP8); - - vis_and(TMP8, MASK_fe, TMP8); - - vis_and(DST_0, REF_0, TMP10); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP8, TMP8); - - vis_and(DST_2, REF_2, TMP12); - vis_ld64_2(dest, stride_8, DST_2); - - vis_ld64(ref[0], TMP14); - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - - dest += stride; - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_2); - - vis_xor(DST_0, REF_0, TMP20); - - vis_and(TMP20, MASK_fe, TMP20); - - vis_xor(DST_2, REF_2, TMP22); - vis_mul8x16(CONST_128, TMP20, TMP20); - - vis_and(TMP22, MASK_fe, TMP22); - - vis_and(DST_0, REF_0, TMP24); - vis_mul8x16(CONST_128, TMP22, TMP22); - - vis_and(DST_2, REF_2, TMP26); - - vis_and(TMP20, MASK_7f, TMP20); - - vis_and(TMP22, MASK_7f, TMP22); - - vis_padd16(TMP24, TMP20, TMP20); - vis_st64(TMP20, dest[0]); - - vis_padd16(TMP26, TMP22, TMP22); - vis_st64_2(TMP22, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(dest[0], DST_0); + + vis_ld64(dest[8], DST_2); + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP2, TMP4, REF_2); + + vis_ld64(constants128[0], CONST_128); + + ref += stride; + height = (height >> 1) - 1; + + do { /* 24 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP6, MASK_fe, TMP6); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_xor(DST_2, REF_2, TMP8); + + vis_and(TMP8, MASK_fe, TMP8); + + vis_and(DST_0, REF_0, TMP10); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP8, TMP8); + + vis_and(DST_2, REF_2, TMP12); + vis_ld64_2(dest, stride_8, DST_2); + + vis_ld64(ref[0], TMP14); + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + + dest += stride; + vis_ld64_2(ref, 8, TMP16); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 16, TMP18); + vis_faligndata(TMP2, TMP4, REF_2); + ref += stride; + + vis_xor(DST_0, REF_0, TMP20); + + vis_and(TMP20, MASK_fe, TMP20); + + vis_xor(DST_2, REF_2, TMP22); + vis_mul8x16(CONST_128, TMP20, TMP20); + + vis_and(TMP22, MASK_fe, TMP22); + + vis_and(DST_0, REF_0, TMP24); + vis_mul8x16(CONST_128, TMP22, TMP22); + + vis_and(DST_2, REF_2, TMP26); + + vis_ld64_2(dest, stride, DST_0); + vis_faligndata(TMP14, TMP16, REF_0); + + vis_ld64_2(dest, stride_8, DST_2); + vis_faligndata(TMP16, TMP18, REF_2); + + vis_and(TMP20, MASK_7f, TMP20); + + vis_and(TMP22, MASK_7f, TMP22); + + vis_padd16(TMP24, TMP20, TMP20); + vis_st64(TMP20, dest[0]); + + vis_padd16(TMP26, TMP22, TMP22); + vis_st64_2(TMP22, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP6, MASK_fe, TMP6); + + vis_ld64_2(ref, 16, TMP4); + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_xor(DST_2, REF_2, TMP8); + + vis_and(TMP8, MASK_fe, TMP8); + + vis_and(DST_0, REF_0, TMP10); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP8, TMP8); + + vis_and(DST_2, REF_2, TMP12); + vis_ld64_2(dest, stride_8, DST_2); + + vis_ld64(ref[0], TMP14); + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + + dest += stride; + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_2); + + vis_xor(DST_0, REF_0, TMP20); + + vis_and(TMP20, MASK_fe, TMP20); + + vis_xor(DST_2, REF_2, TMP22); + vis_mul8x16(CONST_128, TMP20, TMP20); + + vis_and(TMP22, MASK_fe, TMP22); + + vis_and(DST_0, REF_0, TMP24); + vis_mul8x16(CONST_128, TMP22, TMP22); + + vis_and(DST_2, REF_2, TMP26); + + vis_and(TMP20, MASK_7f, TMP20); + + vis_and(TMP22, MASK_7f, TMP22); + + vis_padd16(TMP24, TMP20, TMP20); + vis_st64(TMP20, dest[0]); + + vis_padd16(TMP26, TMP22, TMP22); + vis_st64_2(TMP22, dest, 8); } static void MC_avg_no_round_o_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(dest[0], DST_0); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants128[0], CONST_128); - - ref += stride; - height = (height >> 1) - 1; - - do { /* 12 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP4); - - vis_ld64(ref[8], TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_and(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - ref += stride; - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_ld64(ref[0], TMP12); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(ref[8], TMP2); - vis_xor(DST_0, REF_0, TMP0); - ref += stride; - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_padd16(TMP6, TMP4, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_and(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - - vis_faligndata(TMP12, TMP2, REF_0); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_padd16(TMP6, TMP0, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(DST_0, REF_0, TMP4); - - vis_ld64(ref[8], TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_and(DST_0, REF_0, TMP6); - vis_ld64_2(dest, stride, DST_0); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_xor(DST_0, REF_0, TMP0); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_padd16(TMP6, TMP4, TMP4); - vis_st64(TMP4, dest[0]); - dest += stride; - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_and(DST_0, REF_0, TMP6); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_padd16(TMP6, TMP0, TMP4); - vis_st64(TMP4, dest[0]); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(dest[0], DST_0); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants128[0], CONST_128); + + ref += stride; + height = (height >> 1) - 1; + + do { /* 12 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP4); + + vis_ld64(ref[8], TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_and(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + ref += stride; + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_ld64(ref[0], TMP12); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(ref[8], TMP2); + vis_xor(DST_0, REF_0, TMP0); + ref += stride; + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_padd16(TMP6, TMP4, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_and(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + + vis_faligndata(TMP12, TMP2, REF_0); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_padd16(TMP6, TMP0, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(DST_0, REF_0, TMP4); + + vis_ld64(ref[8], TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_and(DST_0, REF_0, TMP6); + vis_ld64_2(dest, stride, DST_0); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_xor(DST_0, REF_0, TMP0); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_padd16(TMP6, TMP4, TMP4); + vis_st64(TMP4, dest[0]); + dest += stride; + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_and(DST_0, REF_0, TMP6); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_padd16(TMP6, TMP0, TMP4); + vis_st64(TMP4, dest[0]); } static void MC_put_no_round_x_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants128[0], CONST_128); - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - ref += stride; - height = (height >> 1) - 1; - - do { /* 34 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP8); - - vis_ld64_2(ref, 16, TMP4); - vis_and(TMP6, MASK_fe, TMP6); - ref += stride; - - vis_ld64(ref[0], TMP14); - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_ld64_2(ref, 8, TMP16); - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_and(REF_0, REF_2, TMP10); - - vis_ld64_2(ref, 16, TMP18); - ref += stride; - vis_and(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - - vis_xor(REF_0, REF_2, TMP6); - - vis_xor(REF_4, REF_6, TMP8); - - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_and(REF_0, REF_2, TMP10); - - vis_and(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP14, TMP16, REF_0); - - vis_faligndata(TMP16, TMP18, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP14, TMP16, REF_2); - vis_faligndata(TMP16, TMP18, REF_6); - } else { - vis_src1(TMP16, REF_2); - vis_src1(TMP18, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP6); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP8); - - vis_ld64_2(ref, 16, TMP4); - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_and(REF_0, REF_2, TMP10); - - vis_and(REF_4, REF_6, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); - dest += stride; - - vis_xor(REF_0, REF_2, TMP6); - - vis_xor(REF_4, REF_6, TMP8); - - vis_and(TMP6, MASK_fe, TMP6); - - vis_mul8x16(CONST_128, TMP6, TMP6); - vis_and(TMP8, MASK_fe, TMP8); - - vis_mul8x16(CONST_128, TMP8, TMP8); - vis_and(REF_0, REF_2, TMP10); - - vis_and(REF_4, REF_6, TMP12); - - vis_and(TMP6, MASK_7f, TMP6); - - vis_and(TMP8, MASK_7f, TMP8); - - vis_padd16(TMP10, TMP6, TMP6); - vis_st64(TMP6, dest[0]); - - vis_padd16(TMP12, TMP8, TMP8); - vis_st64_2(TMP8, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants128[0], CONST_128); + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + ref += stride; + height = (height >> 1) - 1; + + do { /* 34 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP8); + + vis_ld64_2(ref, 16, TMP4); + vis_and(TMP6, MASK_fe, TMP6); + ref += stride; + + vis_ld64(ref[0], TMP14); + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_ld64_2(ref, 8, TMP16); + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_and(REF_0, REF_2, TMP10); + + vis_ld64_2(ref, 16, TMP18); + ref += stride; + vis_and(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + + vis_xor(REF_0, REF_2, TMP6); + + vis_xor(REF_4, REF_6, TMP8); + + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_and(REF_0, REF_2, TMP10); + + vis_and(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP14, TMP16, REF_0); + + vis_faligndata(TMP16, TMP18, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP14, TMP16, REF_2); + vis_faligndata(TMP16, TMP18, REF_6); + } else { + vis_src1(TMP16, REF_2); + vis_src1(TMP18, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP6); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP8); + + vis_ld64_2(ref, 16, TMP4); + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_and(REF_0, REF_2, TMP10); + + vis_and(REF_4, REF_6, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); + dest += stride; + + vis_xor(REF_0, REF_2, TMP6); + + vis_xor(REF_4, REF_6, TMP8); + + vis_and(TMP6, MASK_fe, TMP6); + + vis_mul8x16(CONST_128, TMP6, TMP6); + vis_and(TMP8, MASK_fe, TMP8); + + vis_mul8x16(CONST_128, TMP8, TMP8); + vis_and(REF_0, REF_2, TMP10); + + vis_and(REF_4, REF_6, TMP12); + + vis_and(TMP6, MASK_7f, TMP6); + + vis_and(TMP8, MASK_7f, TMP8); + + vis_padd16(TMP10, TMP6, TMP6); + vis_st64(TMP6, dest[0]); + + vis_padd16(TMP12, TMP8, TMP8); + vis_st64_2(TMP8, dest, 8); } static void MC_put_no_round_x_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_ld64(constants_fe[0], MASK_fe); - - vis_ld64(constants_7f[0], MASK_7f); - - vis_ld64(constants128[0], CONST_128); - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - ref += stride; - height = (height >> 1) - 1; - - do { /* 20 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - ref += stride; - - vis_ld64(ref[0], TMP8); - vis_and(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, 8, TMP10); - ref += stride; - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - vis_and(TMP4, MASK_7f, TMP4); - - vis_padd16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(REF_0, REF_2, TMP14); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_alignaddr_g0((void *)off); - vis_faligndata(TMP8, TMP10, REF_0); - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP8, TMP10, REF_2); - } else { - vis_src1(TMP10, REF_2); - } - - vis_and(TMP12, MASK_7f, TMP12); - - vis_padd16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_and(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_alignaddr_g0((void *)off); - - vis_faligndata(TMP0, TMP2, REF_0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - } else { - vis_src1(TMP2, REF_2); - } - - vis_and(TMP4, MASK_7f, TMP4); - - vis_padd16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(REF_0, REF_2, TMP14); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_padd16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_ld64(constants_fe[0], MASK_fe); + + vis_ld64(constants_7f[0], MASK_7f); + + vis_ld64(constants128[0], CONST_128); + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + ref += stride; + height = (height >> 1) - 1; + + do { /* 20 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + ref += stride; + + vis_ld64(ref[0], TMP8); + vis_and(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, 8, TMP10); + ref += stride; + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + vis_and(TMP4, MASK_7f, TMP4); + + vis_padd16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(REF_0, REF_2, TMP14); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_alignaddr_g0((void *)off); + vis_faligndata(TMP8, TMP10, REF_0); + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP8, TMP10, REF_2); + } else { + vis_src1(TMP10, REF_2); + } + + vis_and(TMP12, MASK_7f, TMP12); + + vis_padd16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_and(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_alignaddr_g0((void *)off); + + vis_faligndata(TMP0, TMP2, REF_0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + } else { + vis_src1(TMP2, REF_2); + } + + vis_and(TMP4, MASK_7f, TMP4); + + vis_padd16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(REF_0, REF_2, TMP14); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_padd16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; } static void MC_avg_no_round_x_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - vis_ld64(constants3[0], CONST_3); - vis_fzero(ZERO); - vis_ld64(constants256_512[0], CONST_256); - - ref = vis_alignaddr(ref); - do { /* 26 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64(ref[8], TMP2); - - vis_alignaddr_g0((void *)off); - - vis_ld64(ref[16], TMP4); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(dest[8], DST_2); - vis_faligndata(TMP2, TMP4, REF_4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - } - - vis_mul8x16au(REF_0, CONST_256, TMP0); - - vis_pmerge(ZERO, REF_2, TMP4); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_mul8x16al(DST_0, CONST_512, TMP4); - vis_padd16(TMP2, TMP6, TMP2); - - vis_mul8x16al(DST_1, CONST_512, TMP6); - - vis_mul8x16au(REF_6, CONST_256, TMP12); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_6_1, CONST_256, TMP14); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4, CONST_256, TMP16); - - vis_padd16(TMP0, CONST_3, TMP8); - vis_mul8x16au(REF_4_1, CONST_256, TMP18); - - vis_padd16(TMP2, CONST_3, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_padd16(TMP16, TMP12, TMP0); - - vis_st64(DST_0, dest[0]); - vis_mul8x16al(DST_2, CONST_512, TMP4); - vis_padd16(TMP18, TMP14, TMP2); - - vis_mul8x16al(DST_3, CONST_512, TMP6); - vis_padd16(TMP0, CONST_3, TMP0); - - vis_padd16(TMP2, CONST_3, TMP2); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_padd16(TMP2, TMP6, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64(DST_2, dest[8]); - - ref += stride; - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + vis_ld64(constants3[0], CONST_3); + vis_fzero(ZERO); + vis_ld64(constants256_512[0], CONST_256); + + ref = vis_alignaddr(ref); + do { /* 26 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64(ref[8], TMP2); + + vis_alignaddr_g0((void *)off); + + vis_ld64(ref[16], TMP4); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(dest[8], DST_2); + vis_faligndata(TMP2, TMP4, REF_4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + } + + vis_mul8x16au(REF_0, CONST_256, TMP0); + + vis_pmerge(ZERO, REF_2, TMP4); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_mul8x16al(DST_0, CONST_512, TMP4); + vis_padd16(TMP2, TMP6, TMP2); + + vis_mul8x16al(DST_1, CONST_512, TMP6); + + vis_mul8x16au(REF_6, CONST_256, TMP12); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_6_1, CONST_256, TMP14); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4, CONST_256, TMP16); + + vis_padd16(TMP0, CONST_3, TMP8); + vis_mul8x16au(REF_4_1, CONST_256, TMP18); + + vis_padd16(TMP2, CONST_3, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_padd16(TMP16, TMP12, TMP0); + + vis_st64(DST_0, dest[0]); + vis_mul8x16al(DST_2, CONST_512, TMP4); + vis_padd16(TMP18, TMP14, TMP2); + + vis_mul8x16al(DST_3, CONST_512, TMP6); + vis_padd16(TMP0, CONST_3, TMP0); + + vis_padd16(TMP2, CONST_3, TMP2); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_padd16(TMP2, TMP6, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64(DST_2, dest[8]); + + ref += stride; + dest += stride; + } while (--height); } static void MC_avg_no_round_x_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_times_2 = stride << 1; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - vis_ld64(constants3[0], CONST_3); - vis_fzero(ZERO); - vis_ld64(constants256_512[0], CONST_256); - - ref = vis_alignaddr(ref); - height >>= 2; - do { /* 47 cycles */ - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - - vis_alignaddr_g0((void *)off); - - vis_ld64(ref[0], TMP4); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 8, TMP6); - ref += stride; - - vis_ld64(ref[0], TMP8); - - vis_ld64_2(ref, 8, TMP10); - ref += stride; - vis_faligndata(TMP4, TMP6, REF_4); - - vis_ld64(ref[0], TMP12); - - vis_ld64_2(ref, 8, TMP14); - ref += stride; - vis_faligndata(TMP8, TMP10, REF_S0); - - vis_faligndata(TMP12, TMP14, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64_2(dest, stride, DST_2); - vis_faligndata(TMP4, TMP6, REF_6); - - vis_faligndata(TMP8, TMP10, REF_S2); - - vis_faligndata(TMP12, TMP14, REF_S6); - } else { - vis_ld64(dest[0], DST_0); - vis_src1(TMP2, REF_2); - - vis_ld64_2(dest, stride, DST_2); - vis_src1(TMP6, REF_6); - - vis_src1(TMP10, REF_S2); - - vis_src1(TMP14, REF_S6); - } - - vis_pmerge(ZERO, REF_0, TMP0); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_pmerge(ZERO, REF_2, TMP4); - vis_mul8x16au(REF_2_1, CONST_256, TMP6); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16al(DST_0, CONST_512, TMP16); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16al(DST_1, CONST_512, TMP18); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_4, CONST_256, TMP8); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4_1, CONST_256, TMP10); - - vis_padd16(TMP0, TMP16, TMP0); - vis_mul8x16au(REF_6, CONST_256, TMP12); - - vis_padd16(TMP2, TMP18, TMP2); - vis_mul8x16au(REF_6_1, CONST_256, TMP14); - - vis_padd16(TMP8, CONST_3, TMP8); - vis_mul8x16al(DST_2, CONST_512, TMP16); - - vis_padd16(TMP8, TMP12, TMP8); - vis_mul8x16al(DST_3, CONST_512, TMP18); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP0, DST_0); - - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP10, CONST_3, TMP10); - - vis_ld64_2(dest, stride, DST_0); - vis_padd16(TMP8, TMP16, TMP8); - - vis_ld64_2(dest, stride_times_2, TMP4/*DST_2*/); - vis_padd16(TMP10, TMP18, TMP10); - vis_pack16(TMP8, DST_2); - - vis_pack16(TMP10, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - - vis_mul8x16au(REF_S0_1, CONST_256, TMP2); - vis_pmerge(ZERO, REF_S0, TMP0); - - vis_pmerge(ZERO, REF_S2, TMP24); - vis_mul8x16au(REF_S2_1, CONST_256, TMP6); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16au(REF_S4, CONST_256, TMP8); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16au(REF_S4_1, CONST_256, TMP10); - - vis_padd16(TMP0, TMP24, TMP0); - vis_mul8x16au(REF_S6, CONST_256, TMP12); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_S6_1, CONST_256, TMP14); - - vis_padd16(TMP8, CONST_3, TMP8); - vis_mul8x16al(DST_0, CONST_512, TMP16); - - vis_padd16(TMP10, CONST_3, TMP10); - vis_mul8x16al(DST_1, CONST_512, TMP18); - - vis_padd16(TMP8, TMP12, TMP8); - vis_mul8x16al(TMP4/*DST_2*/, CONST_512, TMP20); - - vis_mul8x16al(TMP5/*DST_3*/, CONST_512, TMP22); - vis_padd16(TMP0, TMP16, TMP0); - - vis_padd16(TMP2, TMP18, TMP2); - vis_pack16(TMP0, DST_0); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_padd16(TMP8, TMP20, TMP8); - - vis_padd16(TMP10, TMP22, TMP10); - vis_pack16(TMP8, DST_2); - - vis_pack16(TMP10, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_times_2 = stride << 1; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + vis_ld64(constants3[0], CONST_3); + vis_fzero(ZERO); + vis_ld64(constants256_512[0], CONST_256); + + ref = vis_alignaddr(ref); + height >>= 2; + do { /* 47 cycles */ + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + + vis_alignaddr_g0((void *)off); + + vis_ld64(ref[0], TMP4); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 8, TMP6); + ref += stride; + + vis_ld64(ref[0], TMP8); + + vis_ld64_2(ref, 8, TMP10); + ref += stride; + vis_faligndata(TMP4, TMP6, REF_4); + + vis_ld64(ref[0], TMP12); + + vis_ld64_2(ref, 8, TMP14); + ref += stride; + vis_faligndata(TMP8, TMP10, REF_S0); + + vis_faligndata(TMP12, TMP14, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64_2(dest, stride, DST_2); + vis_faligndata(TMP4, TMP6, REF_6); + + vis_faligndata(TMP8, TMP10, REF_S2); + + vis_faligndata(TMP12, TMP14, REF_S6); + } else { + vis_ld64(dest[0], DST_0); + vis_src1(TMP2, REF_2); + + vis_ld64_2(dest, stride, DST_2); + vis_src1(TMP6, REF_6); + + vis_src1(TMP10, REF_S2); + + vis_src1(TMP14, REF_S6); + } + + vis_pmerge(ZERO, REF_0, TMP0); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_pmerge(ZERO, REF_2, TMP4); + vis_mul8x16au(REF_2_1, CONST_256, TMP6); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16al(DST_0, CONST_512, TMP16); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16al(DST_1, CONST_512, TMP18); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_4, CONST_256, TMP8); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4_1, CONST_256, TMP10); + + vis_padd16(TMP0, TMP16, TMP0); + vis_mul8x16au(REF_6, CONST_256, TMP12); + + vis_padd16(TMP2, TMP18, TMP2); + vis_mul8x16au(REF_6_1, CONST_256, TMP14); + + vis_padd16(TMP8, CONST_3, TMP8); + vis_mul8x16al(DST_2, CONST_512, TMP16); + + vis_padd16(TMP8, TMP12, TMP8); + vis_mul8x16al(DST_3, CONST_512, TMP18); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP0, DST_0); + + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP10, CONST_3, TMP10); + + vis_ld64_2(dest, stride, DST_0); + vis_padd16(TMP8, TMP16, TMP8); + + vis_ld64_2(dest, stride_times_2, TMP4/*DST_2*/); + vis_padd16(TMP10, TMP18, TMP10); + vis_pack16(TMP8, DST_2); + + vis_pack16(TMP10, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + + vis_mul8x16au(REF_S0_1, CONST_256, TMP2); + vis_pmerge(ZERO, REF_S0, TMP0); + + vis_pmerge(ZERO, REF_S2, TMP24); + vis_mul8x16au(REF_S2_1, CONST_256, TMP6); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16au(REF_S4, CONST_256, TMP8); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16au(REF_S4_1, CONST_256, TMP10); + + vis_padd16(TMP0, TMP24, TMP0); + vis_mul8x16au(REF_S6, CONST_256, TMP12); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_S6_1, CONST_256, TMP14); + + vis_padd16(TMP8, CONST_3, TMP8); + vis_mul8x16al(DST_0, CONST_512, TMP16); + + vis_padd16(TMP10, CONST_3, TMP10); + vis_mul8x16al(DST_1, CONST_512, TMP18); + + vis_padd16(TMP8, TMP12, TMP8); + vis_mul8x16al(TMP4/*DST_2*/, CONST_512, TMP20); + + vis_mul8x16al(TMP5/*DST_3*/, CONST_512, TMP22); + vis_padd16(TMP0, TMP16, TMP0); + + vis_padd16(TMP2, TMP18, TMP2); + vis_pack16(TMP0, DST_0); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_padd16(TMP8, TMP20, TMP8); + + vis_padd16(TMP10, TMP22, TMP10); + vis_pack16(TMP8, DST_2); + + vis_pack16(TMP10, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_put_no_round_y_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - - vis_ld64(ref[0], TMP6); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 8, TMP8); - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64_2(ref, 16, TMP10); - ref += stride; - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP6, TMP8, REF_2); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP8, TMP10, REF_6); - - vis_ld64(constants128[0], CONST_128); - height = (height >> 1) - 1; - do { /* 24 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP12); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP16); - - vis_ld64_2(ref, 16, TMP4); - ref += stride; - vis_and(REF_0, REF_2, TMP14); - - vis_ld64(ref[0], TMP6); - vis_and(REF_4, REF_6, TMP18); - - vis_ld64_2(ref, 8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, 16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(TMP16, MASK_fe, TMP16); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_mul8x16(CONST_128, TMP16, TMP16); - vis_xor(REF_0, REF_2, TMP0); - - vis_xor(REF_4, REF_6, TMP2); - - vis_and(REF_0, REF_2, TMP20); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_and(TMP16, MASK_7f, TMP16); - - vis_padd16(TMP14, TMP12, TMP12); - vis_st64(TMP12, dest[0]); - - vis_padd16(TMP18, TMP16, TMP16); - vis_st64_2(TMP16, dest, 8); - dest += stride; - - vis_and(REF_4, REF_6, TMP18); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP2, MASK_fe, TMP2); - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_faligndata(TMP6, TMP8, REF_2); - vis_mul8x16(CONST_128, TMP2, TMP2); - - vis_faligndata(TMP8, TMP10, REF_6); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_and(TMP2, MASK_7f, TMP2); - - vis_padd16(TMP20, TMP0, TMP0); - vis_st64(TMP0, dest[0]); - - vis_padd16(TMP18, TMP2, TMP2); - vis_st64_2(TMP2, dest, 8); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP12); - - vis_ld64_2(ref, 8, TMP2); - vis_xor(REF_4, REF_6, TMP16); - - vis_ld64_2(ref, 16, TMP4); - vis_and(REF_0, REF_2, TMP14); - - vis_and(REF_4, REF_6, TMP18); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_faligndata(TMP2, TMP4, REF_4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_and(TMP16, MASK_fe, TMP16); - vis_mul8x16(CONST_128, TMP12, TMP12); - - vis_mul8x16(CONST_128, TMP16, TMP16); - vis_xor(REF_0, REF_2, TMP0); - - vis_xor(REF_4, REF_6, TMP2); - - vis_and(REF_0, REF_2, TMP20); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_and(TMP16, MASK_7f, TMP16); - - vis_padd16(TMP14, TMP12, TMP12); - vis_st64(TMP12, dest[0]); - - vis_padd16(TMP18, TMP16, TMP16); - vis_st64_2(TMP16, dest, 8); - dest += stride; - - vis_and(REF_4, REF_6, TMP18); - - vis_and(TMP0, MASK_fe, TMP0); - - vis_and(TMP2, MASK_fe, TMP2); - vis_mul8x16(CONST_128, TMP0, TMP0); - - vis_mul8x16(CONST_128, TMP2, TMP2); - - vis_and(TMP0, MASK_7f, TMP0); - - vis_and(TMP2, MASK_7f, TMP2); - - vis_padd16(TMP20, TMP0, TMP0); - vis_st64(TMP0, dest[0]); - - vis_padd16(TMP18, TMP2, TMP2); - vis_st64_2(TMP2, dest, 8); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + + vis_ld64(ref[0], TMP6); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 8, TMP8); + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64_2(ref, 16, TMP10); + ref += stride; + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP6, TMP8, REF_2); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP8, TMP10, REF_6); + + vis_ld64(constants128[0], CONST_128); + height = (height >> 1) - 1; + do { /* 24 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP12); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP16); + + vis_ld64_2(ref, 16, TMP4); + ref += stride; + vis_and(REF_0, REF_2, TMP14); + + vis_ld64(ref[0], TMP6); + vis_and(REF_4, REF_6, TMP18); + + vis_ld64_2(ref, 8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, 16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(TMP16, MASK_fe, TMP16); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_mul8x16(CONST_128, TMP16, TMP16); + vis_xor(REF_0, REF_2, TMP0); + + vis_xor(REF_4, REF_6, TMP2); + + vis_and(REF_0, REF_2, TMP20); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_and(TMP16, MASK_7f, TMP16); + + vis_padd16(TMP14, TMP12, TMP12); + vis_st64(TMP12, dest[0]); + + vis_padd16(TMP18, TMP16, TMP16); + vis_st64_2(TMP16, dest, 8); + dest += stride; + + vis_and(REF_4, REF_6, TMP18); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP2, MASK_fe, TMP2); + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_faligndata(TMP6, TMP8, REF_2); + vis_mul8x16(CONST_128, TMP2, TMP2); + + vis_faligndata(TMP8, TMP10, REF_6); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_and(TMP2, MASK_7f, TMP2); + + vis_padd16(TMP20, TMP0, TMP0); + vis_st64(TMP0, dest[0]); + + vis_padd16(TMP18, TMP2, TMP2); + vis_st64_2(TMP2, dest, 8); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP12); + + vis_ld64_2(ref, 8, TMP2); + vis_xor(REF_4, REF_6, TMP16); + + vis_ld64_2(ref, 16, TMP4); + vis_and(REF_0, REF_2, TMP14); + + vis_and(REF_4, REF_6, TMP18); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_faligndata(TMP2, TMP4, REF_4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_and(TMP16, MASK_fe, TMP16); + vis_mul8x16(CONST_128, TMP12, TMP12); + + vis_mul8x16(CONST_128, TMP16, TMP16); + vis_xor(REF_0, REF_2, TMP0); + + vis_xor(REF_4, REF_6, TMP2); + + vis_and(REF_0, REF_2, TMP20); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_and(TMP16, MASK_7f, TMP16); + + vis_padd16(TMP14, TMP12, TMP12); + vis_st64(TMP12, dest[0]); + + vis_padd16(TMP18, TMP16, TMP16); + vis_st64_2(TMP16, dest, 8); + dest += stride; + + vis_and(REF_4, REF_6, TMP18); + + vis_and(TMP0, MASK_fe, TMP0); + + vis_and(TMP2, MASK_fe, TMP2); + vis_mul8x16(CONST_128, TMP0, TMP0); + + vis_mul8x16(CONST_128, TMP2, TMP2); + + vis_and(TMP0, MASK_7f, TMP0); + + vis_and(TMP2, MASK_7f, TMP2); + + vis_padd16(TMP20, TMP0, TMP0); + vis_st64(TMP0, dest[0]); + + vis_padd16(TMP18, TMP2, TMP2); + vis_st64_2(TMP2, dest, 8); } static void MC_put_no_round_y_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - - ref = vis_alignaddr(ref); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - - vis_ld64(ref[0], TMP4); - - vis_ld64_2(ref, 8, TMP6); - ref += stride; - - vis_ld64(constants_fe[0], MASK_fe); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64(constants_7f[0], MASK_7f); - vis_faligndata(TMP4, TMP6, REF_2); - - vis_ld64(constants128[0], CONST_128); - height = (height >> 1) - 1; - do { /* 12 cycles */ - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - vis_and(TMP4, MASK_fe, TMP4); - - vis_and(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - vis_ld64(ref[0], TMP0); - - vis_ld64_2(ref, 8, TMP2); - ref += stride; - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_mul8x16(CONST_128, TMP12, TMP12); - vis_and(REF_0, REF_2, TMP14); - - vis_padd16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_faligndata(TMP0, TMP2, REF_2); - - vis_and(TMP12, MASK_7f, TMP12); - - vis_padd16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - } while (--height); - - vis_ld64(ref[0], TMP0); - vis_xor(REF_0, REF_2, TMP4); - - vis_ld64_2(ref, 8, TMP2); - vis_and(TMP4, MASK_fe, TMP4); - - vis_and(REF_0, REF_2, TMP6); - vis_mul8x16(CONST_128, TMP4, TMP4); - - vis_faligndata(TMP0, TMP2, REF_0); - - vis_xor(REF_0, REF_2, TMP12); - - vis_and(TMP4, MASK_7f, TMP4); - - vis_and(TMP12, MASK_fe, TMP12); - - vis_mul8x16(CONST_128, TMP12, TMP12); - vis_and(REF_0, REF_2, TMP14); - - vis_padd16(TMP6, TMP4, DST_0); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_and(TMP12, MASK_7f, TMP12); - - vis_padd16(TMP14, TMP12, DST_0); - vis_st64(DST_0, dest[0]); + uint8_t *ref = (uint8_t *) _ref; + + ref = vis_alignaddr(ref); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + + vis_ld64(ref[0], TMP4); + + vis_ld64_2(ref, 8, TMP6); + ref += stride; + + vis_ld64(constants_fe[0], MASK_fe); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64(constants_7f[0], MASK_7f); + vis_faligndata(TMP4, TMP6, REF_2); + + vis_ld64(constants128[0], CONST_128); + height = (height >> 1) - 1; + do { /* 12 cycles */ + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + vis_and(TMP4, MASK_fe, TMP4); + + vis_and(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + vis_ld64(ref[0], TMP0); + + vis_ld64_2(ref, 8, TMP2); + ref += stride; + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_mul8x16(CONST_128, TMP12, TMP12); + vis_and(REF_0, REF_2, TMP14); + + vis_padd16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_faligndata(TMP0, TMP2, REF_2); + + vis_and(TMP12, MASK_7f, TMP12); + + vis_padd16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + } while (--height); + + vis_ld64(ref[0], TMP0); + vis_xor(REF_0, REF_2, TMP4); + + vis_ld64_2(ref, 8, TMP2); + vis_and(TMP4, MASK_fe, TMP4); + + vis_and(REF_0, REF_2, TMP6); + vis_mul8x16(CONST_128, TMP4, TMP4); + + vis_faligndata(TMP0, TMP2, REF_0); + + vis_xor(REF_0, REF_2, TMP12); + + vis_and(TMP4, MASK_7f, TMP4); + + vis_and(TMP12, MASK_fe, TMP12); + + vis_mul8x16(CONST_128, TMP12, TMP12); + vis_and(REF_0, REF_2, TMP14); + + vis_padd16(TMP6, TMP4, DST_0); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_and(TMP12, MASK_7f, TMP12); + + vis_padd16(TMP14, TMP12, DST_0); + vis_st64(DST_0, dest[0]); } static void MC_avg_no_round_y_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants3[0], CONST_3); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_6); - height >>= 1; - - do { /* 31 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_pmerge(ZERO, REF_2, TMP12); - vis_mul8x16au(REF_2_1, CONST_256, TMP14); - - vis_ld64_2(ref, stride_8, TMP2); - vis_pmerge(ZERO, REF_6, TMP16); - vis_mul8x16au(REF_6_1, CONST_256, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(dest, 8, DST_2); - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64_2(ref, stride, TMP6); - vis_pmerge(ZERO, REF_0, TMP0); - vis_mul8x16au(REF_0_1, CONST_256, TMP2); - - vis_ld64_2(ref, stride_8, TMP8); - vis_pmerge(ZERO, REF_4, TMP4); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - - vis_ld64_2(dest, stride, REF_S0/*DST_4*/); - vis_faligndata(TMP6, TMP8, REF_2); - vis_mul8x16au(REF_4_1, CONST_256, TMP6); - - vis_ld64_2(dest, stride_8, REF_S2/*DST_6*/); - vis_faligndata(TMP8, TMP10, REF_6); - vis_mul8x16al(DST_0, CONST_512, TMP20); - - vis_padd16(TMP0, CONST_3, TMP0); - vis_mul8x16al(DST_1, CONST_512, TMP22); - - vis_padd16(TMP2, CONST_3, TMP2); - vis_mul8x16al(DST_2, CONST_512, TMP24); - - vis_padd16(TMP4, CONST_3, TMP4); - vis_mul8x16al(DST_3, CONST_512, TMP26); - - vis_padd16(TMP6, CONST_3, TMP6); - - vis_padd16(TMP12, TMP20, TMP12); - vis_mul8x16al(REF_S0, CONST_512, TMP20); - - vis_padd16(TMP14, TMP22, TMP14); - vis_mul8x16al(REF_S0_1, CONST_512, TMP22); - - vis_padd16(TMP16, TMP24, TMP16); - vis_mul8x16al(REF_S2, CONST_512, TMP24); - - vis_padd16(TMP18, TMP26, TMP18); - vis_mul8x16al(REF_S2_1, CONST_512, TMP26); - - vis_padd16(TMP12, TMP0, TMP12); - vis_mul8x16au(REF_2, CONST_256, TMP28); - - vis_padd16(TMP14, TMP2, TMP14); - vis_mul8x16au(REF_2_1, CONST_256, TMP30); - - vis_padd16(TMP16, TMP4, TMP16); - vis_mul8x16au(REF_6, CONST_256, REF_S4); - - vis_padd16(TMP18, TMP6, TMP18); - vis_mul8x16au(REF_6_1, CONST_256, REF_S6); - - vis_pack16(TMP12, DST_0); - vis_padd16(TMP28, TMP0, TMP12); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP30, TMP2, TMP14); - - vis_pack16(TMP16, DST_2); - vis_padd16(REF_S4, TMP4, TMP16); - - vis_pack16(TMP18, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - vis_padd16(REF_S6, TMP6, TMP18); - - vis_padd16(TMP12, TMP20, TMP12); - - vis_padd16(TMP14, TMP22, TMP14); - vis_pack16(TMP12, DST_0); - - vis_padd16(TMP16, TMP24, TMP16); - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - - vis_padd16(TMP18, TMP26, TMP18); - vis_pack16(TMP16, DST_2); - - vis_pack16(TMP18, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants3[0], CONST_3); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_6); + height >>= 1; + + do { /* 31 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_pmerge(ZERO, REF_2, TMP12); + vis_mul8x16au(REF_2_1, CONST_256, TMP14); + + vis_ld64_2(ref, stride_8, TMP2); + vis_pmerge(ZERO, REF_6, TMP16); + vis_mul8x16au(REF_6_1, CONST_256, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(dest, 8, DST_2); + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64_2(ref, stride, TMP6); + vis_pmerge(ZERO, REF_0, TMP0); + vis_mul8x16au(REF_0_1, CONST_256, TMP2); + + vis_ld64_2(ref, stride_8, TMP8); + vis_pmerge(ZERO, REF_4, TMP4); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + + vis_ld64_2(dest, stride, REF_S0/*DST_4*/); + vis_faligndata(TMP6, TMP8, REF_2); + vis_mul8x16au(REF_4_1, CONST_256, TMP6); + + vis_ld64_2(dest, stride_8, REF_S2/*DST_6*/); + vis_faligndata(TMP8, TMP10, REF_6); + vis_mul8x16al(DST_0, CONST_512, TMP20); + + vis_padd16(TMP0, CONST_3, TMP0); + vis_mul8x16al(DST_1, CONST_512, TMP22); + + vis_padd16(TMP2, CONST_3, TMP2); + vis_mul8x16al(DST_2, CONST_512, TMP24); + + vis_padd16(TMP4, CONST_3, TMP4); + vis_mul8x16al(DST_3, CONST_512, TMP26); + + vis_padd16(TMP6, CONST_3, TMP6); + + vis_padd16(TMP12, TMP20, TMP12); + vis_mul8x16al(REF_S0, CONST_512, TMP20); + + vis_padd16(TMP14, TMP22, TMP14); + vis_mul8x16al(REF_S0_1, CONST_512, TMP22); + + vis_padd16(TMP16, TMP24, TMP16); + vis_mul8x16al(REF_S2, CONST_512, TMP24); + + vis_padd16(TMP18, TMP26, TMP18); + vis_mul8x16al(REF_S2_1, CONST_512, TMP26); + + vis_padd16(TMP12, TMP0, TMP12); + vis_mul8x16au(REF_2, CONST_256, TMP28); + + vis_padd16(TMP14, TMP2, TMP14); + vis_mul8x16au(REF_2_1, CONST_256, TMP30); + + vis_padd16(TMP16, TMP4, TMP16); + vis_mul8x16au(REF_6, CONST_256, REF_S4); + + vis_padd16(TMP18, TMP6, TMP18); + vis_mul8x16au(REF_6_1, CONST_256, REF_S6); + + vis_pack16(TMP12, DST_0); + vis_padd16(TMP28, TMP0, TMP12); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP30, TMP2, TMP14); + + vis_pack16(TMP16, DST_2); + vis_padd16(REF_S4, TMP4, TMP16); + + vis_pack16(TMP18, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + vis_padd16(REF_S6, TMP6, TMP18); + + vis_padd16(TMP12, TMP20, TMP12); + + vis_padd16(TMP14, TMP22, TMP14); + vis_pack16(TMP12, DST_0); + + vis_padd16(TMP16, TMP24, TMP16); + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + + vis_padd16(TMP18, TMP26, TMP18); + vis_pack16(TMP16, DST_2); + + vis_pack16(TMP18, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_avg_no_round_y_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - int stride_8 = stride + 8; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(constants3[0], CONST_3); - vis_faligndata(TMP0, TMP2, REF_2); - - vis_ld64(constants256_512[0], CONST_256); - - height >>= 1; - do { /* 20 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_pmerge(ZERO, REF_2, TMP8); - vis_mul8x16au(REF_2_1, CONST_256, TMP10); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - - vis_ld64(dest[0], DST_0); - - vis_ld64_2(dest, stride, DST_2); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride, TMP4); - vis_mul8x16al(DST_0, CONST_512, TMP16); - vis_pmerge(ZERO, REF_0, TMP12); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - vis_mul8x16al(DST_1, CONST_512, TMP18); - vis_pmerge(ZERO, REF_0_1, TMP14); - - vis_padd16(TMP12, CONST_3, TMP12); - vis_mul8x16al(DST_2, CONST_512, TMP24); - - vis_padd16(TMP14, CONST_3, TMP14); - vis_mul8x16al(DST_3, CONST_512, TMP26); - - vis_faligndata(TMP4, TMP6, REF_2); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - vis_mul8x16au(REF_2, CONST_256, TMP20); - - vis_padd16(TMP8, TMP16, TMP0); - vis_mul8x16au(REF_2_1, CONST_256, TMP22); - - vis_padd16(TMP10, TMP18, TMP2); - vis_pack16(TMP0, DST_0); - - vis_pack16(TMP2, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP12, TMP20, TMP12); - - vis_padd16(TMP14, TMP22, TMP14); - - vis_padd16(TMP12, TMP24, TMP0); - - vis_padd16(TMP14, TMP26, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + int stride_8 = stride + 8; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(constants3[0], CONST_3); + vis_faligndata(TMP0, TMP2, REF_2); + + vis_ld64(constants256_512[0], CONST_256); + + height >>= 1; + do { /* 20 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_pmerge(ZERO, REF_2, TMP8); + vis_mul8x16au(REF_2_1, CONST_256, TMP10); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + + vis_ld64(dest[0], DST_0); + + vis_ld64_2(dest, stride, DST_2); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride, TMP4); + vis_mul8x16al(DST_0, CONST_512, TMP16); + vis_pmerge(ZERO, REF_0, TMP12); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + vis_mul8x16al(DST_1, CONST_512, TMP18); + vis_pmerge(ZERO, REF_0_1, TMP14); + + vis_padd16(TMP12, CONST_3, TMP12); + vis_mul8x16al(DST_2, CONST_512, TMP24); + + vis_padd16(TMP14, CONST_3, TMP14); + vis_mul8x16al(DST_3, CONST_512, TMP26); + + vis_faligndata(TMP4, TMP6, REF_2); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + vis_mul8x16au(REF_2, CONST_256, TMP20); + + vis_padd16(TMP8, TMP16, TMP0); + vis_mul8x16au(REF_2_1, CONST_256, TMP22); + + vis_padd16(TMP10, TMP18, TMP2); + vis_pack16(TMP0, DST_0); + + vis_pack16(TMP2, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP12, TMP20, TMP12); + + vis_padd16(TMP14, TMP22, TMP14); + + vis_padd16(TMP12, TMP24, TMP0); + + vis_padd16(TMP14, TMP26, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_put_no_round_xy_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants1[0], CONST_1); - vis_faligndata(TMP0, TMP2, REF_S0); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - vis_faligndata(TMP2, TMP4, REF_S6); - } else { - vis_src1(TMP2, REF_S2); - vis_src1(TMP4, REF_S6); - } - - height >>= 1; - do { - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S0_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - vis_mul8x16au(REF_S2, CONST_256, TMP16); - vis_pmerge(ZERO, REF_S2_1, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - vis_mul8x16au(REF_S4, CONST_256, TMP20); - vis_pmerge(ZERO, REF_S4_1, TMP22); - - vis_ld64_2(ref, stride, TMP6); - vis_mul8x16au(REF_S6, CONST_256, TMP24); - vis_pmerge(ZERO, REF_S6_1, TMP26); - - vis_ld64_2(ref, stride_8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_faligndata(TMP6, TMP8, REF_S0); - - vis_faligndata(TMP8, TMP10, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - vis_faligndata(TMP6, TMP8, REF_S2); - vis_faligndata(TMP8, TMP10, REF_S6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - vis_src1(TMP8, REF_S2); - vis_src1(TMP10, REF_S6); - } - - vis_mul8x16au(REF_0, CONST_256, TMP0); - vis_pmerge(ZERO, REF_0_1, TMP2); - - vis_mul8x16au(REF_2, CONST_256, TMP4); - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_padd16(TMP0, CONST_2, TMP8); - vis_mul8x16au(REF_4, CONST_256, TMP0); - - vis_padd16(TMP2, CONST_1, TMP10); - vis_mul8x16au(REF_4_1, CONST_256, TMP2); - - vis_padd16(TMP8, TMP4, TMP8); - vis_mul8x16au(REF_6, CONST_256, TMP4); - - vis_padd16(TMP10, TMP6, TMP10); - vis_mul8x16au(REF_6_1, CONST_256, TMP6); - - vis_padd16(TMP12, TMP8, TMP12); - - vis_padd16(TMP14, TMP10, TMP14); - - vis_padd16(TMP12, TMP16, TMP12); - - vis_padd16(TMP14, TMP18, TMP14); - vis_pack16(TMP12, DST_0); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP0, CONST_1, TMP12); - - vis_mul8x16au(REF_S0, CONST_256, TMP0); - vis_padd16(TMP2, CONST_1, TMP14); - - vis_mul8x16au(REF_S0_1, CONST_256, TMP2); - vis_padd16(TMP12, TMP4, TMP12); - - vis_mul8x16au(REF_S2, CONST_256, TMP4); - vis_padd16(TMP14, TMP6, TMP14); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP6); - vis_padd16(TMP20, TMP12, TMP20); - - vis_padd16(TMP22, TMP14, TMP22); - - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP22, TMP26, TMP22); - vis_pack16(TMP20, DST_2); - - vis_pack16(TMP22, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - vis_padd16(TMP0, TMP4, TMP24); - - vis_mul8x16au(REF_S4, CONST_256, TMP0); - vis_padd16(TMP2, TMP6, TMP26); - - vis_mul8x16au(REF_S4_1, CONST_256, TMP2); - vis_padd16(TMP24, TMP8, TMP24); - - vis_padd16(TMP26, TMP10, TMP26); - vis_pack16(TMP24, DST_0); - - vis_pack16(TMP26, DST_1); - vis_st64(DST_0, dest[0]); - vis_pmerge(ZERO, REF_S6, TMP4); - - vis_pmerge(ZERO, REF_S6_1, TMP6); - - vis_padd16(TMP0, TMP4, TMP0); - - vis_padd16(TMP2, TMP6, TMP2); - - vis_padd16(TMP0, TMP12, TMP0); - - vis_padd16(TMP2, TMP14, TMP2); - vis_pack16(TMP0, DST_2); - - vis_pack16(TMP2, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants1[0], CONST_1); + vis_faligndata(TMP0, TMP2, REF_S0); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + vis_faligndata(TMP2, TMP4, REF_S6); + } else { + vis_src1(TMP2, REF_S2); + vis_src1(TMP4, REF_S6); + } + + height >>= 1; + do { + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S0_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + vis_mul8x16au(REF_S2, CONST_256, TMP16); + vis_pmerge(ZERO, REF_S2_1, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + vis_mul8x16au(REF_S4, CONST_256, TMP20); + vis_pmerge(ZERO, REF_S4_1, TMP22); + + vis_ld64_2(ref, stride, TMP6); + vis_mul8x16au(REF_S6, CONST_256, TMP24); + vis_pmerge(ZERO, REF_S6_1, TMP26); + + vis_ld64_2(ref, stride_8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_faligndata(TMP6, TMP8, REF_S0); + + vis_faligndata(TMP8, TMP10, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + vis_faligndata(TMP6, TMP8, REF_S2); + vis_faligndata(TMP8, TMP10, REF_S6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + vis_src1(TMP8, REF_S2); + vis_src1(TMP10, REF_S6); + } + + vis_mul8x16au(REF_0, CONST_256, TMP0); + vis_pmerge(ZERO, REF_0_1, TMP2); + + vis_mul8x16au(REF_2, CONST_256, TMP4); + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_padd16(TMP0, CONST_2, TMP8); + vis_mul8x16au(REF_4, CONST_256, TMP0); + + vis_padd16(TMP2, CONST_1, TMP10); + vis_mul8x16au(REF_4_1, CONST_256, TMP2); + + vis_padd16(TMP8, TMP4, TMP8); + vis_mul8x16au(REF_6, CONST_256, TMP4); + + vis_padd16(TMP10, TMP6, TMP10); + vis_mul8x16au(REF_6_1, CONST_256, TMP6); + + vis_padd16(TMP12, TMP8, TMP12); + + vis_padd16(TMP14, TMP10, TMP14); + + vis_padd16(TMP12, TMP16, TMP12); + + vis_padd16(TMP14, TMP18, TMP14); + vis_pack16(TMP12, DST_0); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP0, CONST_1, TMP12); + + vis_mul8x16au(REF_S0, CONST_256, TMP0); + vis_padd16(TMP2, CONST_1, TMP14); + + vis_mul8x16au(REF_S0_1, CONST_256, TMP2); + vis_padd16(TMP12, TMP4, TMP12); + + vis_mul8x16au(REF_S2, CONST_256, TMP4); + vis_padd16(TMP14, TMP6, TMP14); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP6); + vis_padd16(TMP20, TMP12, TMP20); + + vis_padd16(TMP22, TMP14, TMP22); + + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP22, TMP26, TMP22); + vis_pack16(TMP20, DST_2); + + vis_pack16(TMP22, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + vis_padd16(TMP0, TMP4, TMP24); + + vis_mul8x16au(REF_S4, CONST_256, TMP0); + vis_padd16(TMP2, TMP6, TMP26); + + vis_mul8x16au(REF_S4_1, CONST_256, TMP2); + vis_padd16(TMP24, TMP8, TMP24); + + vis_padd16(TMP26, TMP10, TMP26); + vis_pack16(TMP24, DST_0); + + vis_pack16(TMP26, DST_1); + vis_st64(DST_0, dest[0]); + vis_pmerge(ZERO, REF_S6, TMP4); + + vis_pmerge(ZERO, REF_S6_1, TMP6); + + vis_padd16(TMP0, TMP4, TMP0); + + vis_padd16(TMP2, TMP6, TMP2); + + vis_padd16(TMP0, TMP12, TMP0); + + vis_padd16(TMP2, TMP14, TMP2); + vis_pack16(TMP0, DST_2); + + vis_pack16(TMP2, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_put_no_round_xy_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - - vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(constants1[0], CONST_1); - - vis_ld64(constants256_512[0], CONST_256); - vis_faligndata(TMP0, TMP2, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - } else { - vis_src1(TMP2, REF_S2); - } - - height >>= 1; - do { /* 26 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP8); - vis_pmerge(ZERO, REF_S2, TMP12); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - vis_mul8x16au(REF_S0_1, CONST_256, TMP10); - vis_pmerge(ZERO, REF_S2_1, TMP14); - - vis_ld64_2(ref, stride, TMP4); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - vis_faligndata(TMP0, TMP2, REF_S4); - - vis_pmerge(ZERO, REF_S4, TMP18); - - vis_pmerge(ZERO, REF_S4_1, TMP20); - - vis_faligndata(TMP4, TMP6, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S6); - vis_faligndata(TMP4, TMP6, REF_S2); - } else { - vis_src1(TMP2, REF_S6); - vis_src1(TMP6, REF_S2); - } - - vis_padd16(TMP18, CONST_1, TMP18); - vis_mul8x16au(REF_S6, CONST_256, TMP22); - - vis_padd16(TMP20, CONST_1, TMP20); - vis_mul8x16au(REF_S6_1, CONST_256, TMP24); - - vis_mul8x16au(REF_S0, CONST_256, TMP26); - vis_pmerge(ZERO, REF_S0_1, TMP28); - - vis_mul8x16au(REF_S2, CONST_256, TMP30); - vis_padd16(TMP18, TMP22, TMP18); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP32); - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP8, TMP18, TMP8); - - vis_padd16(TMP10, TMP20, TMP10); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - vis_padd16(TMP18, TMP26, TMP18); - - vis_padd16(TMP20, TMP28, TMP20); - - vis_padd16(TMP18, TMP30, TMP18); - - vis_padd16(TMP20, TMP32, TMP20); - vis_pack16(TMP18, DST_2); - - vis_pack16(TMP20, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + + vis_set_gsr(5 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(constants1[0], CONST_1); + + vis_ld64(constants256_512[0], CONST_256); + vis_faligndata(TMP0, TMP2, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + } else { + vis_src1(TMP2, REF_S2); + } + + height >>= 1; + do { /* 26 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP8); + vis_pmerge(ZERO, REF_S2, TMP12); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + vis_mul8x16au(REF_S0_1, CONST_256, TMP10); + vis_pmerge(ZERO, REF_S2_1, TMP14); + + vis_ld64_2(ref, stride, TMP4); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + vis_faligndata(TMP0, TMP2, REF_S4); + + vis_pmerge(ZERO, REF_S4, TMP18); + + vis_pmerge(ZERO, REF_S4_1, TMP20); + + vis_faligndata(TMP4, TMP6, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S6); + vis_faligndata(TMP4, TMP6, REF_S2); + } else { + vis_src1(TMP2, REF_S6); + vis_src1(TMP6, REF_S2); + } + + vis_padd16(TMP18, CONST_1, TMP18); + vis_mul8x16au(REF_S6, CONST_256, TMP22); + + vis_padd16(TMP20, CONST_1, TMP20); + vis_mul8x16au(REF_S6_1, CONST_256, TMP24); + + vis_mul8x16au(REF_S0, CONST_256, TMP26); + vis_pmerge(ZERO, REF_S0_1, TMP28); + + vis_mul8x16au(REF_S2, CONST_256, TMP30); + vis_padd16(TMP18, TMP22, TMP18); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP32); + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP8, TMP18, TMP8); + + vis_padd16(TMP10, TMP20, TMP10); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + vis_padd16(TMP18, TMP26, TMP18); + + vis_padd16(TMP20, TMP28, TMP20); + + vis_padd16(TMP18, TMP30, TMP18); + + vis_padd16(TMP20, TMP32, TMP20); + vis_pack16(TMP18, DST_2); + + vis_pack16(TMP20, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } static void MC_avg_no_round_xy_16_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - int stride_16 = stride + 16; - - vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[ 0], TMP0); - vis_fzero(ZERO); - - vis_ld64(ref[ 8], TMP2); - - vis_ld64(ref[16], TMP4); - - vis_ld64(constants6[0], CONST_6); - vis_faligndata(TMP0, TMP2, REF_S0); - - vis_ld64(constants256_1024[0], CONST_256); - vis_faligndata(TMP2, TMP4, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - vis_faligndata(TMP2, TMP4, REF_S6); - } else { - vis_src1(TMP2, REF_S2); - vis_src1(TMP4, REF_S6); - } - - height >>= 1; - do { /* 55 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S0_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride_8, TMP2); - vis_mul8x16au(REF_S2, CONST_256, TMP16); - vis_pmerge(ZERO, REF_S2_1, TMP18); - - vis_ld64_2(ref, stride_16, TMP4); - ref += stride; - vis_mul8x16au(REF_S4, CONST_256, TMP20); - vis_pmerge(ZERO, REF_S4_1, TMP22); - - vis_ld64_2(ref, stride, TMP6); - vis_mul8x16au(REF_S6, CONST_256, TMP24); - vis_pmerge(ZERO, REF_S6_1, TMP26); - - vis_ld64_2(ref, stride_8, TMP8); - vis_faligndata(TMP0, TMP2, REF_0); - - vis_ld64_2(ref, stride_16, TMP10); - ref += stride; - vis_faligndata(TMP2, TMP4, REF_4); - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP6, TMP8, REF_S0); - - vis_ld64_2(dest, 8, DST_2); - vis_faligndata(TMP8, TMP10, REF_S4); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_2); - vis_faligndata(TMP2, TMP4, REF_6); - vis_faligndata(TMP6, TMP8, REF_S2); - vis_faligndata(TMP8, TMP10, REF_S6); - } else { - vis_src1(TMP2, REF_2); - vis_src1(TMP4, REF_6); - vis_src1(TMP8, REF_S2); - vis_src1(TMP10, REF_S6); - } - - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_0, TMP0); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_pmerge(ZERO, REF_0_1, TMP2); - - vis_mul8x16au(REF_2, CONST_256, TMP4); - vis_pmerge(ZERO, REF_2_1, TMP6); - - vis_mul8x16al(DST_2, CONST_1024, REF_0); - vis_padd16(TMP0, CONST_6, TMP0); - - vis_mul8x16al(DST_3, CONST_1024, REF_2); - vis_padd16(TMP2, CONST_6, TMP2); - - vis_padd16(TMP0, TMP4, TMP0); - vis_mul8x16au(REF_4, CONST_256, TMP4); - - vis_padd16(TMP2, TMP6, TMP2); - vis_mul8x16au(REF_4_1, CONST_256, TMP6); - - vis_padd16(TMP12, TMP0, TMP12); - vis_mul8x16au(REF_6, CONST_256, TMP8); - - vis_padd16(TMP14, TMP2, TMP14); - vis_mul8x16au(REF_6_1, CONST_256, TMP10); - - vis_padd16(TMP12, TMP16, TMP12); - vis_mul8x16au(REF_S0, CONST_256, REF_4); - - vis_padd16(TMP14, TMP18, TMP14); - vis_mul8x16au(REF_S0_1, CONST_256, REF_6); - - vis_padd16(TMP12, TMP30, TMP12); - - vis_padd16(TMP14, TMP32, TMP14); - vis_pack16(TMP12, DST_0); - - vis_pack16(TMP14, DST_1); - vis_st64(DST_0, dest[0]); - vis_padd16(TMP4, CONST_6, TMP4); - - vis_ld64_2(dest, stride, DST_0); - vis_padd16(TMP6, CONST_6, TMP6); - vis_mul8x16au(REF_S2, CONST_256, TMP12); - - vis_padd16(TMP4, TMP8, TMP4); - vis_mul8x16au(REF_S2_1, CONST_256, TMP14); - - vis_padd16(TMP6, TMP10, TMP6); - - vis_padd16(TMP20, TMP4, TMP20); - - vis_padd16(TMP22, TMP6, TMP22); - - vis_padd16(TMP20, TMP24, TMP20); - - vis_padd16(TMP22, TMP26, TMP22); - - vis_padd16(TMP20, REF_0, TMP20); - vis_mul8x16au(REF_S4, CONST_256, REF_0); - - vis_padd16(TMP22, REF_2, TMP22); - vis_pack16(TMP20, DST_2); - - vis_pack16(TMP22, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - - vis_ld64_2(dest, 8, DST_2); - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_S4_1, REF_2); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_padd16(REF_4, TMP0, TMP8); - - vis_mul8x16au(REF_S6, CONST_256, REF_4); - vis_padd16(REF_6, TMP2, TMP10); - - vis_mul8x16au(REF_S6_1, CONST_256, REF_6); - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - - vis_padd16(TMP8, TMP30, TMP8); - - vis_padd16(TMP10, TMP32, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - - vis_padd16(REF_0, TMP4, REF_0); - - vis_mul8x16al(DST_2, CONST_1024, TMP30); - vis_padd16(REF_2, TMP6, REF_2); - - vis_mul8x16al(DST_3, CONST_1024, TMP32); - vis_padd16(REF_0, REF_4, REF_0); - - vis_padd16(REF_2, REF_6, REF_2); - - vis_padd16(REF_0, TMP30, REF_0); - - /* stall */ - - vis_padd16(REF_2, TMP32, REF_2); - vis_pack16(REF_0, DST_2); - - vis_pack16(REF_2, DST_3); - vis_st64_2(DST_2, dest, 8); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + int stride_16 = stride + 16; + + vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[ 0], TMP0); + vis_fzero(ZERO); + + vis_ld64(ref[ 8], TMP2); + + vis_ld64(ref[16], TMP4); + + vis_ld64(constants6[0], CONST_6); + vis_faligndata(TMP0, TMP2, REF_S0); + + vis_ld64(constants256_1024[0], CONST_256); + vis_faligndata(TMP2, TMP4, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + vis_faligndata(TMP2, TMP4, REF_S6); + } else { + vis_src1(TMP2, REF_S2); + vis_src1(TMP4, REF_S6); + } + + height >>= 1; + do { /* 55 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S0_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride_8, TMP2); + vis_mul8x16au(REF_S2, CONST_256, TMP16); + vis_pmerge(ZERO, REF_S2_1, TMP18); + + vis_ld64_2(ref, stride_16, TMP4); + ref += stride; + vis_mul8x16au(REF_S4, CONST_256, TMP20); + vis_pmerge(ZERO, REF_S4_1, TMP22); + + vis_ld64_2(ref, stride, TMP6); + vis_mul8x16au(REF_S6, CONST_256, TMP24); + vis_pmerge(ZERO, REF_S6_1, TMP26); + + vis_ld64_2(ref, stride_8, TMP8); + vis_faligndata(TMP0, TMP2, REF_0); + + vis_ld64_2(ref, stride_16, TMP10); + ref += stride; + vis_faligndata(TMP2, TMP4, REF_4); + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP6, TMP8, REF_S0); + + vis_ld64_2(dest, 8, DST_2); + vis_faligndata(TMP8, TMP10, REF_S4); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_2); + vis_faligndata(TMP2, TMP4, REF_6); + vis_faligndata(TMP6, TMP8, REF_S2); + vis_faligndata(TMP8, TMP10, REF_S6); + } else { + vis_src1(TMP2, REF_2); + vis_src1(TMP4, REF_6); + vis_src1(TMP8, REF_S2); + vis_src1(TMP10, REF_S6); + } + + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_0, TMP0); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_pmerge(ZERO, REF_0_1, TMP2); + + vis_mul8x16au(REF_2, CONST_256, TMP4); + vis_pmerge(ZERO, REF_2_1, TMP6); + + vis_mul8x16al(DST_2, CONST_1024, REF_0); + vis_padd16(TMP0, CONST_6, TMP0); + + vis_mul8x16al(DST_3, CONST_1024, REF_2); + vis_padd16(TMP2, CONST_6, TMP2); + + vis_padd16(TMP0, TMP4, TMP0); + vis_mul8x16au(REF_4, CONST_256, TMP4); + + vis_padd16(TMP2, TMP6, TMP2); + vis_mul8x16au(REF_4_1, CONST_256, TMP6); + + vis_padd16(TMP12, TMP0, TMP12); + vis_mul8x16au(REF_6, CONST_256, TMP8); + + vis_padd16(TMP14, TMP2, TMP14); + vis_mul8x16au(REF_6_1, CONST_256, TMP10); + + vis_padd16(TMP12, TMP16, TMP12); + vis_mul8x16au(REF_S0, CONST_256, REF_4); + + vis_padd16(TMP14, TMP18, TMP14); + vis_mul8x16au(REF_S0_1, CONST_256, REF_6); + + vis_padd16(TMP12, TMP30, TMP12); + + vis_padd16(TMP14, TMP32, TMP14); + vis_pack16(TMP12, DST_0); + + vis_pack16(TMP14, DST_1); + vis_st64(DST_0, dest[0]); + vis_padd16(TMP4, CONST_6, TMP4); + + vis_ld64_2(dest, stride, DST_0); + vis_padd16(TMP6, CONST_6, TMP6); + vis_mul8x16au(REF_S2, CONST_256, TMP12); + + vis_padd16(TMP4, TMP8, TMP4); + vis_mul8x16au(REF_S2_1, CONST_256, TMP14); + + vis_padd16(TMP6, TMP10, TMP6); + + vis_padd16(TMP20, TMP4, TMP20); + + vis_padd16(TMP22, TMP6, TMP22); + + vis_padd16(TMP20, TMP24, TMP20); + + vis_padd16(TMP22, TMP26, TMP22); + + vis_padd16(TMP20, REF_0, TMP20); + vis_mul8x16au(REF_S4, CONST_256, REF_0); + + vis_padd16(TMP22, REF_2, TMP22); + vis_pack16(TMP20, DST_2); + + vis_pack16(TMP22, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + + vis_ld64_2(dest, 8, DST_2); + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_S4_1, REF_2); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_padd16(REF_4, TMP0, TMP8); + + vis_mul8x16au(REF_S6, CONST_256, REF_4); + vis_padd16(REF_6, TMP2, TMP10); + + vis_mul8x16au(REF_S6_1, CONST_256, REF_6); + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + + vis_padd16(TMP8, TMP30, TMP8); + + vis_padd16(TMP10, TMP32, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + + vis_padd16(REF_0, TMP4, REF_0); + + vis_mul8x16al(DST_2, CONST_1024, TMP30); + vis_padd16(REF_2, TMP6, REF_2); + + vis_mul8x16al(DST_3, CONST_1024, TMP32); + vis_padd16(REF_0, REF_4, REF_0); + + vis_padd16(REF_2, REF_6, REF_2); + + vis_padd16(REF_0, TMP30, REF_0); + + /* stall */ + + vis_padd16(REF_2, TMP32, REF_2); + vis_pack16(REF_0, DST_2); + + vis_pack16(REF_2, DST_3); + vis_st64_2(DST_2, dest, 8); + dest += stride; + } while (--height); } static void MC_avg_no_round_xy_8_vis (uint8_t * dest, const uint8_t * _ref, - const int stride, int height) + const int stride, int height) { - uint8_t *ref = (uint8_t *) _ref; - unsigned long off = (unsigned long) ref & 0x7; - unsigned long off_plus_1 = off + 1; - int stride_8 = stride + 8; - - vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); - - ref = vis_alignaddr(ref); - - vis_ld64(ref[0], TMP0); - vis_fzero(ZERO); - - vis_ld64_2(ref, 8, TMP2); - - vis_ld64(constants6[0], CONST_6); - - vis_ld64(constants256_1024[0], CONST_256); - vis_faligndata(TMP0, TMP2, REF_S0); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S2); - } else { - vis_src1(TMP2, REF_S2); - } - - height >>= 1; - do { /* 31 cycles */ - vis_ld64_2(ref, stride, TMP0); - vis_mul8x16au(REF_S0, CONST_256, TMP8); - vis_pmerge(ZERO, REF_S0_1, TMP10); - - vis_ld64_2(ref, stride_8, TMP2); - ref += stride; - vis_mul8x16au(REF_S2, CONST_256, TMP12); - vis_pmerge(ZERO, REF_S2_1, TMP14); - - vis_alignaddr_g0((void *)off); - - vis_ld64_2(ref, stride, TMP4); - vis_faligndata(TMP0, TMP2, REF_S4); - - vis_ld64_2(ref, stride_8, TMP6); - ref += stride; - - vis_ld64(dest[0], DST_0); - vis_faligndata(TMP4, TMP6, REF_S0); - - vis_ld64_2(dest, stride, DST_2); - - if (off != 0x7) { - vis_alignaddr_g0((void *)off_plus_1); - vis_faligndata(TMP0, TMP2, REF_S6); - vis_faligndata(TMP4, TMP6, REF_S2); - } else { - vis_src1(TMP2, REF_S6); - vis_src1(TMP6, REF_S2); - } - - vis_mul8x16al(DST_0, CONST_1024, TMP30); - vis_pmerge(ZERO, REF_S4, TMP22); - - vis_mul8x16al(DST_1, CONST_1024, TMP32); - vis_pmerge(ZERO, REF_S4_1, TMP24); - - vis_mul8x16au(REF_S6, CONST_256, TMP26); - vis_pmerge(ZERO, REF_S6_1, TMP28); - - vis_mul8x16au(REF_S0, CONST_256, REF_S4); - vis_padd16(TMP22, CONST_6, TMP22); - - vis_mul8x16au(REF_S0_1, CONST_256, REF_S6); - vis_padd16(TMP24, CONST_6, TMP24); - - vis_mul8x16al(DST_2, CONST_1024, REF_0); - vis_padd16(TMP22, TMP26, TMP22); - - vis_mul8x16al(DST_3, CONST_1024, REF_2); - vis_padd16(TMP24, TMP28, TMP24); - - vis_mul8x16au(REF_S2, CONST_256, TMP26); - vis_padd16(TMP8, TMP22, TMP8); - - vis_mul8x16au(REF_S2_1, CONST_256, TMP28); - vis_padd16(TMP10, TMP24, TMP10); - - vis_padd16(TMP8, TMP12, TMP8); - - vis_padd16(TMP10, TMP14, TMP10); - - vis_padd16(TMP8, TMP30, TMP8); - - vis_padd16(TMP10, TMP32, TMP10); - vis_pack16(TMP8, DST_0); - - vis_pack16(TMP10, DST_1); - vis_st64(DST_0, dest[0]); - dest += stride; - - vis_padd16(REF_S4, TMP22, TMP12); - - vis_padd16(REF_S6, TMP24, TMP14); - - vis_padd16(TMP12, TMP26, TMP12); - - vis_padd16(TMP14, TMP28, TMP14); - - vis_padd16(TMP12, REF_0, TMP12); - - vis_padd16(TMP14, REF_2, TMP14); - vis_pack16(TMP12, DST_2); - - vis_pack16(TMP14, DST_3); - vis_st64(DST_2, dest[0]); - dest += stride; - } while (--height); + uint8_t *ref = (uint8_t *) _ref; + unsigned long off = (unsigned long) ref & 0x7; + unsigned long off_plus_1 = off + 1; + int stride_8 = stride + 8; + + vis_set_gsr(4 << VIS_GSR_SCALEFACT_SHIFT); + + ref = vis_alignaddr(ref); + + vis_ld64(ref[0], TMP0); + vis_fzero(ZERO); + + vis_ld64_2(ref, 8, TMP2); + + vis_ld64(constants6[0], CONST_6); + + vis_ld64(constants256_1024[0], CONST_256); + vis_faligndata(TMP0, TMP2, REF_S0); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S2); + } else { + vis_src1(TMP2, REF_S2); + } + + height >>= 1; + do { /* 31 cycles */ + vis_ld64_2(ref, stride, TMP0); + vis_mul8x16au(REF_S0, CONST_256, TMP8); + vis_pmerge(ZERO, REF_S0_1, TMP10); + + vis_ld64_2(ref, stride_8, TMP2); + ref += stride; + vis_mul8x16au(REF_S2, CONST_256, TMP12); + vis_pmerge(ZERO, REF_S2_1, TMP14); + + vis_alignaddr_g0((void *)off); + + vis_ld64_2(ref, stride, TMP4); + vis_faligndata(TMP0, TMP2, REF_S4); + + vis_ld64_2(ref, stride_8, TMP6); + ref += stride; + + vis_ld64(dest[0], DST_0); + vis_faligndata(TMP4, TMP6, REF_S0); + + vis_ld64_2(dest, stride, DST_2); + + if (off != 0x7) { + vis_alignaddr_g0((void *)off_plus_1); + vis_faligndata(TMP0, TMP2, REF_S6); + vis_faligndata(TMP4, TMP6, REF_S2); + } else { + vis_src1(TMP2, REF_S6); + vis_src1(TMP6, REF_S2); + } + + vis_mul8x16al(DST_0, CONST_1024, TMP30); + vis_pmerge(ZERO, REF_S4, TMP22); + + vis_mul8x16al(DST_1, CONST_1024, TMP32); + vis_pmerge(ZERO, REF_S4_1, TMP24); + + vis_mul8x16au(REF_S6, CONST_256, TMP26); + vis_pmerge(ZERO, REF_S6_1, TMP28); + + vis_mul8x16au(REF_S0, CONST_256, REF_S4); + vis_padd16(TMP22, CONST_6, TMP22); + + vis_mul8x16au(REF_S0_1, CONST_256, REF_S6); + vis_padd16(TMP24, CONST_6, TMP24); + + vis_mul8x16al(DST_2, CONST_1024, REF_0); + vis_padd16(TMP22, TMP26, TMP22); + + vis_mul8x16al(DST_3, CONST_1024, REF_2); + vis_padd16(TMP24, TMP28, TMP24); + + vis_mul8x16au(REF_S2, CONST_256, TMP26); + vis_padd16(TMP8, TMP22, TMP8); + + vis_mul8x16au(REF_S2_1, CONST_256, TMP28); + vis_padd16(TMP10, TMP24, TMP10); + + vis_padd16(TMP8, TMP12, TMP8); + + vis_padd16(TMP10, TMP14, TMP10); + + vis_padd16(TMP8, TMP30, TMP8); + + vis_padd16(TMP10, TMP32, TMP10); + vis_pack16(TMP8, DST_0); + + vis_pack16(TMP10, DST_1); + vis_st64(DST_0, dest[0]); + dest += stride; + + vis_padd16(REF_S4, TMP22, TMP12); + + vis_padd16(REF_S6, TMP24, TMP14); + + vis_padd16(TMP12, TMP26, TMP12); + + vis_padd16(TMP14, TMP28, TMP14); + + vis_padd16(TMP12, REF_0, TMP12); + + vis_padd16(TMP14, REF_2, TMP14); + vis_pack16(TMP12, DST_2); + + vis_pack16(TMP14, DST_3); + vis_st64(DST_2, dest[0]); + dest += stride; + } while (--height); } /* End of no rounding code */
--- a/sparc/vis.h Wed Dec 21 17:50:40 2005 +0000 +++ b/sparc/vis.h Thu Dec 22 01:10:11 2005 +0000 @@ -41,151 +41,151 @@ * the assembler to keep the binary from becoming tainted. */ -#define vis_opc_base ((0x1 << 31) | (0x36 << 19)) -#define vis_opf(X) ((X) << 5) -#define vis_sreg(X) (X) -#define vis_dreg(X) (((X)&0x1f)|((X)>>5)) -#define vis_rs1_s(X) (vis_sreg(X) << 14) -#define vis_rs1_d(X) (vis_dreg(X) << 14) -#define vis_rs2_s(X) (vis_sreg(X) << 0) -#define vis_rs2_d(X) (vis_dreg(X) << 0) -#define vis_rd_s(X) (vis_sreg(X) << 25) -#define vis_rd_d(X) (vis_dreg(X) << 25) +#define vis_opc_base ((0x1 << 31) | (0x36 << 19)) +#define vis_opf(X) ((X) << 5) +#define vis_sreg(X) (X) +#define vis_dreg(X) (((X)&0x1f)|((X)>>5)) +#define vis_rs1_s(X) (vis_sreg(X) << 14) +#define vis_rs1_d(X) (vis_dreg(X) << 14) +#define vis_rs2_s(X) (vis_sreg(X) << 0) +#define vis_rs2_d(X) (vis_dreg(X) << 0) +#define vis_rd_s(X) (vis_sreg(X) << 25) +#define vis_rd_d(X) (vis_dreg(X) << 25) #define vis_ss2s(opf,rs1,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_s(rs1) | \ vis_rs2_s(rs2) | \ vis_rd_s(rd))) #define vis_dd2d(opf,rs1,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_d(rs1) | \ vis_rs2_d(rs2) | \ vis_rd_d(rd))) #define vis_ss2d(opf,rs1,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_s(rs1) | \ vis_rs2_s(rs2) | \ vis_rd_d(rd))) #define vis_sd2d(opf,rs1,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_s(rs1) | \ vis_rs2_d(rs2) | \ vis_rd_d(rd))) #define vis_d2s(opf,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs2_d(rs2) | \ vis_rd_s(rd))) #define vis_s2d(opf,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs2_s(rs2) | \ vis_rd_d(rd))) #define vis_d12d(opf,rs1,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_d(rs1) | \ vis_rd_d(rd))) #define vis_d22d(opf,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs2_d(rs2) | \ vis_rd_d(rd))) #define vis_s12s(opf,rs1,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs1_s(rs1) | \ vis_rd_s(rd))) #define vis_s22s(opf,rs2,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rs2_s(rs2) | \ vis_rd_s(rd))) #define vis_s(opf,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rd_s(rd))) #define vis_d(opf,rd) \ - __asm__ __volatile__ (".word %0" \ - : : "i" (vis_opc_base | vis_opf(opf) | \ + __asm__ __volatile__ (".word %0" \ + : : "i" (vis_opc_base | vis_opf(opf) | \ vis_rd_d(rd))) #define vis_r2m(op,rd,mem) \ - __asm__ __volatile__ (#op "\t%%f" #rd ", [%0]" : : "r" (&(mem)) ) + __asm__ __volatile__ (#op "\t%%f" #rd ", [%0]" : : "r" (&(mem)) ) #define vis_r2m_2(op,rd,mem1,mem2) \ - __asm__ __volatile__ (#op "\t%%f" #rd ", [%0 + %1]" : : "r" (mem1), "r" (mem2) ) + __asm__ __volatile__ (#op "\t%%f" #rd ", [%0 + %1]" : : "r" (mem1), "r" (mem2) ) #define vis_m2r(op,mem,rd) \ - __asm__ __volatile__ (#op "\t[%0], %%f" #rd : : "r" (&(mem)) ) + __asm__ __volatile__ (#op "\t[%0], %%f" #rd : : "r" (&(mem)) ) #define vis_m2r_2(op,mem1,mem2,rd) \ - __asm__ __volatile__ (#op "\t[%0 + %1], %%f" #rd : : "r" (mem1), "r" (mem2) ) + __asm__ __volatile__ (#op "\t[%0 + %1], %%f" #rd : : "r" (mem1), "r" (mem2) ) static inline void vis_set_gsr(unsigned int _val) { - register unsigned int val asm("g1"); + register unsigned int val asm("g1"); - val = _val; - __asm__ __volatile__(".word 0xa7804000" - : : "r" (val)); + val = _val; + __asm__ __volatile__(".word 0xa7804000" + : : "r" (val)); } -#define VIS_GSR_ALIGNADDR_MASK 0x0000007 -#define VIS_GSR_ALIGNADDR_SHIFT 0 -#define VIS_GSR_SCALEFACT_MASK 0x0000078 -#define VIS_GSR_SCALEFACT_SHIFT 3 +#define VIS_GSR_ALIGNADDR_MASK 0x0000007 +#define VIS_GSR_ALIGNADDR_SHIFT 0 +#define VIS_GSR_SCALEFACT_MASK 0x0000078 +#define VIS_GSR_SCALEFACT_SHIFT 3 -#define vis_ld32(mem,rs1) vis_m2r(ld, mem, rs1) -#define vis_ld32_2(mem1,mem2,rs1) vis_m2r_2(ld, mem1, mem2, rs1) -#define vis_st32(rs1,mem) vis_r2m(st, rs1, mem) -#define vis_st32_2(rs1,mem1,mem2) vis_r2m_2(st, rs1, mem1, mem2) -#define vis_ld64(mem,rs1) vis_m2r(ldd, mem, rs1) -#define vis_ld64_2(mem1,mem2,rs1) vis_m2r_2(ldd, mem1, mem2, rs1) -#define vis_st64(rs1,mem) vis_r2m(std, rs1, mem) -#define vis_st64_2(rs1,mem1,mem2) vis_r2m_2(std, rs1, mem1, mem2) +#define vis_ld32(mem,rs1) vis_m2r(ld, mem, rs1) +#define vis_ld32_2(mem1,mem2,rs1) vis_m2r_2(ld, mem1, mem2, rs1) +#define vis_st32(rs1,mem) vis_r2m(st, rs1, mem) +#define vis_st32_2(rs1,mem1,mem2) vis_r2m_2(st, rs1, mem1, mem2) +#define vis_ld64(mem,rs1) vis_m2r(ldd, mem, rs1) +#define vis_ld64_2(mem1,mem2,rs1) vis_m2r_2(ldd, mem1, mem2, rs1) +#define vis_st64(rs1,mem) vis_r2m(std, rs1, mem) +#define vis_st64_2(rs1,mem1,mem2) vis_r2m_2(std, rs1, mem1, mem2) #define vis_ldblk(mem, rd) \ -do { register void *__mem asm("g1"); \ - __mem = &(mem); \ - __asm__ __volatile__(".word 0xc1985e00 | %1" \ - : \ - : "r" (__mem), \ - "i" (vis_rd_d(rd)) \ - : "memory"); \ +do { register void *__mem asm("g1"); \ + __mem = &(mem); \ + __asm__ __volatile__(".word 0xc1985e00 | %1" \ + : \ + : "r" (__mem), \ + "i" (vis_rd_d(rd)) \ + : "memory"); \ } while (0) #define vis_stblk(rd, mem) \ -do { register void *__mem asm("g1"); \ - __mem = &(mem); \ - __asm__ __volatile__(".word 0xc1b85e00 | %1" \ - : \ - : "r" (__mem), \ - "i" (vis_rd_d(rd)) \ - : "memory"); \ +do { register void *__mem asm("g1"); \ + __mem = &(mem); \ + __asm__ __volatile__(".word 0xc1b85e00 | %1" \ + : \ + : "r" (__mem), \ + "i" (vis_rd_d(rd)) \ + : "memory"); \ } while (0) -#define vis_membar_storestore() \ - __asm__ __volatile__(".word 0x8143e008" : : : "memory") +#define vis_membar_storestore() \ + __asm__ __volatile__(".word 0x8143e008" : : : "memory") -#define vis_membar_sync() \ - __asm__ __volatile__(".word 0x8143e040" : : : "memory") +#define vis_membar_sync() \ + __asm__ __volatile__(".word 0x8143e040" : : : "memory") /* 16 and 32 bit partitioned addition and subtraction. The normal * versions perform 4 16-bit or 2 32-bit additions or subtractions. @@ -193,136 +193,136 @@ * subtractions. */ -#define vis_padd16(rs1,rs2,rd) vis_dd2d(0x50, rs1, rs2, rd) -#define vis_padd16s(rs1,rs2,rd) vis_ss2s(0x51, rs1, rs2, rd) -#define vis_padd32(rs1,rs2,rd) vis_dd2d(0x52, rs1, rs2, rd) -#define vis_padd32s(rs1,rs2,rd) vis_ss2s(0x53, rs1, rs2, rd) -#define vis_psub16(rs1,rs2,rd) vis_dd2d(0x54, rs1, rs2, rd) -#define vis_psub16s(rs1,rs2,rd) vis_ss2s(0x55, rs1, rs2, rd) -#define vis_psub32(rs1,rs2,rd) vis_dd2d(0x56, rs1, rs2, rd) -#define vis_psub32s(rs1,rs2,rd) vis_ss2s(0x57, rs1, rs2, rd) +#define vis_padd16(rs1,rs2,rd) vis_dd2d(0x50, rs1, rs2, rd) +#define vis_padd16s(rs1,rs2,rd) vis_ss2s(0x51, rs1, rs2, rd) +#define vis_padd32(rs1,rs2,rd) vis_dd2d(0x52, rs1, rs2, rd) +#define vis_padd32s(rs1,rs2,rd) vis_ss2s(0x53, rs1, rs2, rd) +#define vis_psub16(rs1,rs2,rd) vis_dd2d(0x54, rs1, rs2, rd) +#define vis_psub16s(rs1,rs2,rd) vis_ss2s(0x55, rs1, rs2, rd) +#define vis_psub32(rs1,rs2,rd) vis_dd2d(0x56, rs1, rs2, rd) +#define vis_psub32s(rs1,rs2,rd) vis_ss2s(0x57, rs1, rs2, rd) /* Pixel formatting instructions. */ -#define vis_pack16(rs2,rd) vis_d2s( 0x3b, rs2, rd) -#define vis_pack32(rs1,rs2,rd) vis_dd2d(0x3a, rs1, rs2, rd) -#define vis_packfix(rs2,rd) vis_d2s( 0x3d, rs2, rd) -#define vis_expand(rs2,rd) vis_s2d( 0x4d, rs2, rd) -#define vis_pmerge(rs1,rs2,rd) vis_ss2d(0x4b, rs1, rs2, rd) +#define vis_pack16(rs2,rd) vis_d2s( 0x3b, rs2, rd) +#define vis_pack32(rs1,rs2,rd) vis_dd2d(0x3a, rs1, rs2, rd) +#define vis_packfix(rs2,rd) vis_d2s( 0x3d, rs2, rd) +#define vis_expand(rs2,rd) vis_s2d( 0x4d, rs2, rd) +#define vis_pmerge(rs1,rs2,rd) vis_ss2d(0x4b, rs1, rs2, rd) /* Partitioned multiply instructions. */ -#define vis_mul8x16(rs1,rs2,rd) vis_sd2d(0x31, rs1, rs2, rd) -#define vis_mul8x16au(rs1,rs2,rd) vis_ss2d(0x33, rs1, rs2, rd) -#define vis_mul8x16al(rs1,rs2,rd) vis_ss2d(0x35, rs1, rs2, rd) -#define vis_mul8sux16(rs1,rs2,rd) vis_dd2d(0x36, rs1, rs2, rd) -#define vis_mul8ulx16(rs1,rs2,rd) vis_dd2d(0x37, rs1, rs2, rd) -#define vis_muld8sux16(rs1,rs2,rd) vis_ss2d(0x38, rs1, rs2, rd) -#define vis_muld8ulx16(rs1,rs2,rd) vis_ss2d(0x39, rs1, rs2, rd) +#define vis_mul8x16(rs1,rs2,rd) vis_sd2d(0x31, rs1, rs2, rd) +#define vis_mul8x16au(rs1,rs2,rd) vis_ss2d(0x33, rs1, rs2, rd) +#define vis_mul8x16al(rs1,rs2,rd) vis_ss2d(0x35, rs1, rs2, rd) +#define vis_mul8sux16(rs1,rs2,rd) vis_dd2d(0x36, rs1, rs2, rd) +#define vis_mul8ulx16(rs1,rs2,rd) vis_dd2d(0x37, rs1, rs2, rd) +#define vis_muld8sux16(rs1,rs2,rd) vis_ss2d(0x38, rs1, rs2, rd) +#define vis_muld8ulx16(rs1,rs2,rd) vis_ss2d(0x39, rs1, rs2, rd) /* Alignment instructions. */ static inline void *vis_alignaddr(void *_ptr) { - register void *ptr asm("g1"); + register void *ptr asm("g1"); - ptr = _ptr; + ptr = _ptr; - __asm__ __volatile__(".word %2" - : "=&r" (ptr) - : "0" (ptr), - "i" (vis_opc_base | vis_opf(0x18) | - vis_rs1_s(1) | - vis_rs2_s(0) | - vis_rd_s(1))); + __asm__ __volatile__(".word %2" + : "=&r" (ptr) + : "0" (ptr), + "i" (vis_opc_base | vis_opf(0x18) | + vis_rs1_s(1) | + vis_rs2_s(0) | + vis_rd_s(1))); - return ptr; + return ptr; } static inline void vis_alignaddr_g0(void *_ptr) { - register void *ptr asm("g1"); + register void *ptr asm("g1"); - ptr = _ptr; + ptr = _ptr; - __asm__ __volatile__(".word %2" - : "=&r" (ptr) - : "0" (ptr), - "i" (vis_opc_base | vis_opf(0x18) | - vis_rs1_s(1) | - vis_rs2_s(0) | - vis_rd_s(0))); + __asm__ __volatile__(".word %2" + : "=&r" (ptr) + : "0" (ptr), + "i" (vis_opc_base | vis_opf(0x18) | + vis_rs1_s(1) | + vis_rs2_s(0) | + vis_rd_s(0))); } static inline void *vis_alignaddrl(void *_ptr) { - register void *ptr asm("g1"); + register void *ptr asm("g1"); - ptr = _ptr; + ptr = _ptr; - __asm__ __volatile__(".word %2" - : "=&r" (ptr) - : "0" (ptr), - "i" (vis_opc_base | vis_opf(0x19) | - vis_rs1_s(1) | - vis_rs2_s(0) | - vis_rd_s(1))); + __asm__ __volatile__(".word %2" + : "=&r" (ptr) + : "0" (ptr), + "i" (vis_opc_base | vis_opf(0x19) | + vis_rs1_s(1) | + vis_rs2_s(0) | + vis_rd_s(1))); - return ptr; + return ptr; } static inline void vis_alignaddrl_g0(void *_ptr) { - register void *ptr asm("g1"); + register void *ptr asm("g1"); - ptr = _ptr; + ptr = _ptr; - __asm__ __volatile__(".word %2" - : "=&r" (ptr) - : "0" (ptr), - "i" (vis_opc_base | vis_opf(0x19) | - vis_rs1_s(1) | - vis_rs2_s(0) | - vis_rd_s(0))); + __asm__ __volatile__(".word %2" + : "=&r" (ptr) + : "0" (ptr), + "i" (vis_opc_base | vis_opf(0x19) | + vis_rs1_s(1) | + vis_rs2_s(0) | + vis_rd_s(0))); } -#define vis_faligndata(rs1,rs2,rd) vis_dd2d(0x48, rs1, rs2, rd) +#define vis_faligndata(rs1,rs2,rd) vis_dd2d(0x48, rs1, rs2, rd) /* Logical operate instructions. */ -#define vis_fzero(rd) vis_d( 0x60, rd) -#define vis_fzeros(rd) vis_s( 0x61, rd) -#define vis_fone(rd) vis_d( 0x7e, rd) -#define vis_fones(rd) vis_s( 0x7f, rd) -#define vis_src1(rs1,rd) vis_d12d(0x74, rs1, rd) -#define vis_src1s(rs1,rd) vis_s12s(0x75, rs1, rd) -#define vis_src2(rs2,rd) vis_d22d(0x78, rs2, rd) -#define vis_src2s(rs2,rd) vis_s22s(0x79, rs2, rd) -#define vis_not1(rs1,rd) vis_d12d(0x6a, rs1, rd) -#define vis_not1s(rs1,rd) vis_s12s(0x6b, rs1, rd) -#define vis_not2(rs2,rd) vis_d22d(0x66, rs2, rd) -#define vis_not2s(rs2,rd) vis_s22s(0x67, rs2, rd) -#define vis_or(rs1,rs2,rd) vis_dd2d(0x7c, rs1, rs2, rd) -#define vis_ors(rs1,rs2,rd) vis_ss2s(0x7d, rs1, rs2, rd) -#define vis_nor(rs1,rs2,rd) vis_dd2d(0x62, rs1, rs2, rd) -#define vis_nors(rs1,rs2,rd) vis_ss2s(0x63, rs1, rs2, rd) -#define vis_and(rs1,rs2,rd) vis_dd2d(0x70, rs1, rs2, rd) -#define vis_ands(rs1,rs2,rd) vis_ss2s(0x71, rs1, rs2, rd) -#define vis_nand(rs1,rs2,rd) vis_dd2d(0x6e, rs1, rs2, rd) -#define vis_nands(rs1,rs2,rd) vis_ss2s(0x6f, rs1, rs2, rd) -#define vis_xor(rs1,rs2,rd) vis_dd2d(0x6c, rs1, rs2, rd) -#define vis_xors(rs1,rs2,rd) vis_ss2s(0x6d, rs1, rs2, rd) -#define vis_xnor(rs1,rs2,rd) vis_dd2d(0x72, rs1, rs2, rd) -#define vis_xnors(rs1,rs2,rd) vis_ss2s(0x73, rs1, rs2, rd) -#define vis_ornot1(rs1,rs2,rd) vis_dd2d(0x7a, rs1, rs2, rd) -#define vis_ornot1s(rs1,rs2,rd) vis_ss2s(0x7b, rs1, rs2, rd) -#define vis_ornot2(rs1,rs2,rd) vis_dd2d(0x76, rs1, rs2, rd) -#define vis_ornot2s(rs1,rs2,rd) vis_ss2s(0x77, rs1, rs2, rd) -#define vis_andnot1(rs1,rs2,rd) vis_dd2d(0x68, rs1, rs2, rd) -#define vis_andnot1s(rs1,rs2,rd) vis_ss2s(0x69, rs1, rs2, rd) -#define vis_andnot2(rs1,rs2,rd) vis_dd2d(0x64, rs1, rs2, rd) -#define vis_andnot2s(rs1,rs2,rd) vis_ss2s(0x65, rs1, rs2, rd) +#define vis_fzero(rd) vis_d( 0x60, rd) +#define vis_fzeros(rd) vis_s( 0x61, rd) +#define vis_fone(rd) vis_d( 0x7e, rd) +#define vis_fones(rd) vis_s( 0x7f, rd) +#define vis_src1(rs1,rd) vis_d12d(0x74, rs1, rd) +#define vis_src1s(rs1,rd) vis_s12s(0x75, rs1, rd) +#define vis_src2(rs2,rd) vis_d22d(0x78, rs2, rd) +#define vis_src2s(rs2,rd) vis_s22s(0x79, rs2, rd) +#define vis_not1(rs1,rd) vis_d12d(0x6a, rs1, rd) +#define vis_not1s(rs1,rd) vis_s12s(0x6b, rs1, rd) +#define vis_not2(rs2,rd) vis_d22d(0x66, rs2, rd) +#define vis_not2s(rs2,rd) vis_s22s(0x67, rs2, rd) +#define vis_or(rs1,rs2,rd) vis_dd2d(0x7c, rs1, rs2, rd) +#define vis_ors(rs1,rs2,rd) vis_ss2s(0x7d, rs1, rs2, rd) +#define vis_nor(rs1,rs2,rd) vis_dd2d(0x62, rs1, rs2, rd) +#define vis_nors(rs1,rs2,rd) vis_ss2s(0x63, rs1, rs2, rd) +#define vis_and(rs1,rs2,rd) vis_dd2d(0x70, rs1, rs2, rd) +#define vis_ands(rs1,rs2,rd) vis_ss2s(0x71, rs1, rs2, rd) +#define vis_nand(rs1,rs2,rd) vis_dd2d(0x6e, rs1, rs2, rd) +#define vis_nands(rs1,rs2,rd) vis_ss2s(0x6f, rs1, rs2, rd) +#define vis_xor(rs1,rs2,rd) vis_dd2d(0x6c, rs1, rs2, rd) +#define vis_xors(rs1,rs2,rd) vis_ss2s(0x6d, rs1, rs2, rd) +#define vis_xnor(rs1,rs2,rd) vis_dd2d(0x72, rs1, rs2, rd) +#define vis_xnors(rs1,rs2,rd) vis_ss2s(0x73, rs1, rs2, rd) +#define vis_ornot1(rs1,rs2,rd) vis_dd2d(0x7a, rs1, rs2, rd) +#define vis_ornot1s(rs1,rs2,rd) vis_ss2s(0x7b, rs1, rs2, rd) +#define vis_ornot2(rs1,rs2,rd) vis_dd2d(0x76, rs1, rs2, rd) +#define vis_ornot2s(rs1,rs2,rd) vis_ss2s(0x77, rs1, rs2, rd) +#define vis_andnot1(rs1,rs2,rd) vis_dd2d(0x68, rs1, rs2, rd) +#define vis_andnot1s(rs1,rs2,rd) vis_ss2s(0x69, rs1, rs2, rd) +#define vis_andnot2(rs1,rs2,rd) vis_dd2d(0x64, rs1, rs2, rd) +#define vis_andnot2s(rs1,rs2,rd) vis_ss2s(0x65, rs1, rs2, rd) /* Pixel component distance. */ -#define vis_pdist(rs1,rs2,rd) vis_dd2d(0x3e, rs1, rs2, rd) +#define vis_pdist(rs1,rs2,rd) vis_dd2d(0x3e, rs1, rs2, rd)
--- a/svq1.c Wed Dec 21 17:50:40 2005 +0000 +++ b/svq1.c Thu Dec 22 01:10:11 2005 +0000 @@ -57,10 +57,10 @@ static VLC svq1_intra_mean; static VLC svq1_inter_mean; -#define SVQ1_BLOCK_SKIP 0 -#define SVQ1_BLOCK_INTER 1 -#define SVQ1_BLOCK_INTER_4V 2 -#define SVQ1_BLOCK_INTRA 3 +#define SVQ1_BLOCK_SKIP 0 +#define SVQ1_BLOCK_INTER 1 +#define SVQ1_BLOCK_INTER_4V 2 +#define SVQ1_BLOCK_INTRA 3 typedef struct SVQ1Context { MpegEncContext m; // needed for motion estimation, should not be used for anything else, the idea is to make the motion estimation eventually independant of MpegEncContext, so this will be removed then (FIXME/XXX) @@ -95,8 +95,8 @@ /* motion vector (prediction) */ typedef struct svq1_pmv_s { - int x; - int y; + int x; + int y; } svq1_pmv_t; #include "svq1_cb.h" @@ -176,65 +176,65 @@ for (; level > 0; i++) {\ /* process next depth */\ if (i == m) {\ - m = n;\ - if (--level == 0)\ - break;\ + m = n;\ + if (--level == 0)\ + break;\ }\ /* divide block if next bit set */\ if (get_bits (bitbuf, 1) == 0)\ - break;\ + break;\ /* add child nodes */\ list[n++] = list[i];\ list[n++] = list[i] + (((level & 1) ? pitch : 1) << ((level / 2) + 1));\ } #define SVQ1_ADD_CODEBOOK()\ - /* add codebook entries to vector */\ - for (j=0; j < stages; j++) {\ - n3 = codebook[entries[j]] ^ 0x80808080;\ - n1 += ((n3 & 0xFF00FF00) >> 8);\ - n2 += (n3 & 0x00FF00FF);\ - }\ + /* add codebook entries to vector */\ + for (j=0; j < stages; j++) {\ + n3 = codebook[entries[j]] ^ 0x80808080;\ + n1 += ((n3 & 0xFF00FF00) >> 8);\ + n2 += (n3 & 0x00FF00FF);\ + }\ \ - /* clip to [0..255] */\ - if (n1 & 0xFF00FF00) {\ - n3 = ((( n1 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ - n1 += 0x7F007F00;\ - n1 |= (((~n1 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ - n1 &= (n3 & 0x00FF00FF);\ - }\ + /* clip to [0..255] */\ + if (n1 & 0xFF00FF00) {\ + n3 = ((( n1 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ + n1 += 0x7F007F00;\ + n1 |= (((~n1 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ + n1 &= (n3 & 0x00FF00FF);\ + }\ \ - if (n2 & 0xFF00FF00) {\ - n3 = ((( n2 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ - n2 += 0x7F007F00;\ - n2 |= (((~n2 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ - n2 &= (n3 & 0x00FF00FF);\ - } + if (n2 & 0xFF00FF00) {\ + n3 = ((( n2 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ + n2 += 0x7F007F00;\ + n2 |= (((~n2 >> 15) & 0x00010001) | 0x01000100) - 0x00010001;\ + n2 &= (n3 & 0x00FF00FF);\ + } #define SVQ1_DO_CODEBOOK_INTRA()\ for (y=0; y < height; y++) {\ - for (x=0; x < (width / 4); x++, codebook++) {\ - n1 = n4;\ - n2 = n4;\ - SVQ1_ADD_CODEBOOK()\ - /* store result */\ - dst[x] = (n1 << 8) | n2;\ - }\ - dst += (pitch / 4);\ + for (x=0; x < (width / 4); x++, codebook++) {\ + n1 = n4;\ + n2 = n4;\ + SVQ1_ADD_CODEBOOK()\ + /* store result */\ + dst[x] = (n1 << 8) | n2;\ + }\ + dst += (pitch / 4);\ } #define SVQ1_DO_CODEBOOK_NONINTRA()\ for (y=0; y < height; y++) {\ - for (x=0; x < (width / 4); x++, codebook++) {\ - n3 = dst[x];\ - /* add mean value to vector */\ - n1 = ((n3 & 0xFF00FF00) >> 8) + n4;\ - n2 = (n3 & 0x00FF00FF) + n4;\ - SVQ1_ADD_CODEBOOK()\ - /* store result */\ - dst[x] = (n1 << 8) | n2;\ - }\ - dst += (pitch / 4);\ + for (x=0; x < (width / 4); x++, codebook++) {\ + n3 = dst[x];\ + /* add mean value to vector */\ + n1 = ((n3 & 0xFF00FF00) >> 8) + n4;\ + n2 = (n3 & 0x00FF00FF) + n4;\ + SVQ1_ADD_CODEBOOK()\ + /* store result */\ + dst[x] = (n1 << 8) | n2;\ + }\ + dst += (pitch / 4);\ } #define SVQ1_CALC_CODEBOOK_ENTRIES(cbook)\ @@ -242,7 +242,7 @@ bit_cache = get_bits (bitbuf, 4*stages);\ /* calculate codebook entries for this vector */\ for (j=0; j < stages; j++) {\ - entries[j] = (((bit_cache >> (4*(stages - j - 1))) & 0xF) + 16*j) << (level + 1);\ + entries[j] = (((bit_cache >> (4*(stages - j - 1))) & 0xF) + 16*j) << (level + 1);\ }\ mean -= (stages * 128);\ n4 = ((mean + (mean >> 31)) << 16) | (mean & 0xFFFF); @@ -252,9 +252,9 @@ uint8_t *list[63]; uint32_t *dst; const uint32_t *codebook; - int entries[6]; - int i, j, m, n; - int mean, stages; + int entries[6]; + int i, j, m, n; + int mean, stages; unsigned x, y, width, height, level; uint32_t n1, n2, n3, n4; @@ -274,24 +274,24 @@ stages = get_vlc2(bitbuf, svq1_intra_multistage[level].table, 3, 3) - 1; if (stages == -1) { - for (y=0; y < height; y++) { - memset (&dst[y*(pitch / 4)], 0, width); - } - continue; /* skip vector */ + for (y=0; y < height; y++) { + memset (&dst[y*(pitch / 4)], 0, width); + } + continue; /* skip vector */ } if ((stages > 0) && (level >= 4)) { #ifdef DEBUG_SVQ1 av_log(s->avctx, AV_LOG_INFO, "Error (svq1_decode_block_intra): invalid vector: stages=%i level=%i\n",stages,level); #endif - return -1; /* invalid vector */ + return -1; /* invalid vector */ } mean = get_vlc2(bitbuf, svq1_intra_mean.table, 8, 3); if (stages == 0) { for (y=0; y < height; y++) { - memset (&dst[y*(pitch / 4)], mean, width); + memset (&dst[y*(pitch / 4)], mean, width); } } else { SVQ1_CALC_CODEBOOK_ENTRIES(svq1_intra_codebooks); @@ -307,10 +307,10 @@ uint8_t *list[63]; uint32_t *dst; const uint32_t *codebook; - int entries[6]; - int i, j, m, n; - int mean, stages; - int x, y, width, height, level; + int entries[6]; + int i, j, m, n; + int mean, stages; + int x, y, width, height, level; uint32_t n1, n2, n3, n4; /* initialize list for breadth first processing of vectors */ @@ -328,13 +328,13 @@ /* get number of stages (-1 skips vector, 0 for mean only) */ stages = get_vlc2(bitbuf, svq1_inter_multistage[level].table, 3, 2) - 1; - if (stages == -1) continue; /* skip vector */ + if (stages == -1) continue; /* skip vector */ if ((stages > 0) && (level >= 4)) { #ifdef DEBUG_SVQ1 av_log(s->avctx, AV_LOG_INFO, "Error (svq1_decode_block_non_intra): invalid vector: stages=%i level=%i\n",stages,level); #endif - return -1; /* invalid vector */ + return -1; /* invalid vector */ } mean = get_vlc2(bitbuf, svq1_inter_mean.table, 9, 3) - 256; @@ -346,8 +346,8 @@ } static int svq1_decode_motion_vector (GetBitContext *bitbuf, svq1_pmv_t *mv, svq1_pmv_t **pmv) { - int diff; - int i; + int diff; + int i; for (i=0; i < 2; i++) { @@ -372,7 +372,7 @@ static void svq1_skip_block (uint8_t *current, uint8_t *previous, int pitch, int x, int y) { uint8_t *src; uint8_t *dst; - int i; + int i; src = &previous[x + y*pitch]; dst = current; @@ -385,13 +385,13 @@ } static int svq1_motion_inter_block (MpegEncContext *s, GetBitContext *bitbuf, - uint8_t *current, uint8_t *previous, int pitch, - svq1_pmv_t *motion, int x, int y) { + uint8_t *current, uint8_t *previous, int pitch, + svq1_pmv_t *motion, int x, int y) { uint8_t *src; uint8_t *dst; svq1_pmv_t mv; svq1_pmv_t *pmv[3]; - int result; + int result; /* predict and decode motion vector */ pmv[0] = &motion[0]; @@ -409,12 +409,12 @@ if (result != 0) return result; - motion[0].x = - motion[(x / 8) + 2].x = - motion[(x / 8) + 3].x = mv.x; - motion[0].y = - motion[(x / 8) + 2].y = - motion[(x / 8) + 3].y = mv.y; + motion[0].x = + motion[(x / 8) + 2].x = + motion[(x / 8) + 3].x = mv.x; + motion[0].y = + motion[(x / 8) + 2].y = + motion[(x / 8) + 3].y = mv.y; if(y + (mv.y >> 1)<0) mv.y= 0; @@ -437,13 +437,13 @@ } static int svq1_motion_inter_4v_block (MpegEncContext *s, GetBitContext *bitbuf, - uint8_t *current, uint8_t *previous, int pitch, - svq1_pmv_t *motion,int x, int y) { + uint8_t *current, uint8_t *previous, int pitch, + svq1_pmv_t *motion,int x, int y) { uint8_t *src; uint8_t *dst; svq1_pmv_t mv; svq1_pmv_t *pmv[4]; - int i, result; + int i, result; /* predict and decode motion vector (0) */ pmv[0] = &motion[0]; @@ -527,18 +527,18 @@ } static int svq1_decode_delta_block (MpegEncContext *s, GetBitContext *bitbuf, - uint8_t *current, uint8_t *previous, int pitch, - svq1_pmv_t *motion, int x, int y) { + uint8_t *current, uint8_t *previous, int pitch, + svq1_pmv_t *motion, int x, int y) { uint32_t block_type; - int result = 0; + int result = 0; /* get block type */ block_type = get_vlc2(bitbuf, svq1_block_type.table, 2, 2); /* reset motion vectors */ if (block_type == SVQ1_BLOCK_SKIP || block_type == SVQ1_BLOCK_INTRA) { - motion[0].x = - motion[0].y = + motion[0].x = + motion[0].y = motion[(x / 8) + 2].x = motion[(x / 8) + 2].y = motion[(x / 8) + 3].x = @@ -711,8 +711,8 @@ uint8_t *buf, int buf_size) { MpegEncContext *s=avctx->priv_data; - uint8_t *current, *previous; - int result, i, x, y, width, height; + uint8_t *current, *previous; + int result, i, x, y, width, height; AVFrame *pict = data; /* initialize bit buffer */ @@ -781,17 +781,17 @@ if (s->pict_type == I_TYPE) { /* keyframe */ for (y=0; y < height; y+=16) { - for (x=0; x < width; x+=16) { - result = svq1_decode_block_intra (&s->gb, ¤t[x], linesize); - if (result != 0) - { + for (x=0; x < width; x+=16) { + result = svq1_decode_block_intra (&s->gb, ¤t[x], linesize); + if (result != 0) + { //#ifdef DEBUG_SVQ1 - av_log(s->avctx, AV_LOG_INFO, "Error in svq1_decode_block %i (keyframe)\n",result); + av_log(s->avctx, AV_LOG_INFO, "Error in svq1_decode_block %i (keyframe)\n",result); //#endif - return result; - } - } - current += 16*linesize; + return result; + } + } + current += 16*linesize; } } else { svq1_pmv_t pmv[width/8+3]; @@ -799,22 +799,22 @@ memset (pmv, 0, ((width / 8) + 3) * sizeof(svq1_pmv_t)); for (y=0; y < height; y+=16) { - for (x=0; x < width; x+=16) { - result = svq1_decode_delta_block (s, &s->gb, ¤t[x], previous, - linesize, pmv, x, y); - if (result != 0) - { + for (x=0; x < width; x+=16) { + result = svq1_decode_delta_block (s, &s->gb, ¤t[x], previous, + linesize, pmv, x, y); + if (result != 0) + { #ifdef DEBUG_SVQ1 av_log(s->avctx, AV_LOG_INFO, "Error in svq1_decode_delta_block %i\n",result); #endif - return result; - } - } + return result; + } + } - pmv[0].x = - pmv[0].y = 0; + pmv[0].x = + pmv[0].y = 0; - current += 16*linesize; + current += 16*linesize; } } } @@ -902,22 +902,22 @@ /* output 5 unknown bits (2 + 2 + 1) */ put_bits(&s->pb, 5, 0); - for (i = 0; i < 7; i++) - { - if ((svq1_frame_size_table[i].width == s->frame_width) && - (svq1_frame_size_table[i].height == s->frame_height)) - { - put_bits(&s->pb, 3, i); - break; - } - } + for (i = 0; i < 7; i++) + { + if ((svq1_frame_size_table[i].width == s->frame_width) && + (svq1_frame_size_table[i].height == s->frame_height)) + { + put_bits(&s->pb, 3, i); + break; + } + } - if (i == 7) - { - put_bits(&s->pb, 3, 7); - put_bits(&s->pb, 12, s->frame_width); - put_bits(&s->pb, 12, s->frame_height); - } + if (i == 7) + { + put_bits(&s->pb, 3, 7); + put_bits(&s->pb, 12, s->frame_width); + put_bits(&s->pb, 12, s->frame_height); + } } /* no checksum or extra data (next 2 bits get 0) */
--- a/svq3.c Wed Dec 21 17:50:40 2005 +0000 +++ b/svq3.c Thu Dec 22 01:10:11 2005 +0000 @@ -207,7 +207,7 @@ } static inline int svq3_decode_block (GetBitContext *gb, DCTELEM *block, - int index, const int type) { + int index, const int type) { static const uint8_t *const scan_patterns[4] = { luma_dc_zigzag_scan, zigzag_scan, svq3_scan, chroma_dc_scan }; @@ -220,37 +220,37 @@ for (; (vlc = svq3_get_ue_golomb (gb)) != 0; index++) { if (vlc == INVALID_VLC) - return -1; + return -1; sign = (vlc & 0x1) - 1; vlc = (vlc + 1) >> 1; if (type == 3) { - if (vlc < 3) { - run = 0; - level = vlc; - } else if (vlc < 4) { - run = 1; - level = 1; - } else { - run = (vlc & 0x3); - level = ((vlc + 9) >> 2) - run; - } + if (vlc < 3) { + run = 0; + level = vlc; + } else if (vlc < 4) { + run = 1; + level = 1; + } else { + run = (vlc & 0x3); + level = ((vlc + 9) >> 2) - run; + } } else { - if (vlc < 16) { - run = svq3_dct_tables[intra][vlc].run; - level = svq3_dct_tables[intra][vlc].level; - } else if (intra) { - run = (vlc & 0x7); - level = (vlc >> 3) + ((run == 0) ? 8 : ((run < 2) ? 2 : ((run < 5) ? 0 : -1))); - } else { - run = (vlc & 0xF); - level = (vlc >> 4) + ((run == 0) ? 4 : ((run < 3) ? 2 : ((run < 10) ? 1 : 0))); - } + if (vlc < 16) { + run = svq3_dct_tables[intra][vlc].run; + level = svq3_dct_tables[intra][vlc].level; + } else if (intra) { + run = (vlc & 0x7); + level = (vlc >> 3) + ((run == 0) ? 8 : ((run < 2) ? 2 : ((run < 5) ? 0 : -1))); + } else { + run = (vlc & 0xF); + level = (vlc >> 4) + ((run == 0) ? 4 : ((run < 3) ? 2 : ((run < 10) ? 1 : 0))); + } } if ((index += run) >= limit) - return -1; + return -1; block[scan[index]] = (level ^ sign) - sign; } @@ -264,9 +264,9 @@ } static inline void svq3_mc_dir_part (MpegEncContext *s, - int x, int y, int width, int height, - int mx, int my, int dxy, - int thirdpel, int dir, int avg) { + int x, int y, int width, int height, + int mx, int my, int dxy, + int thirdpel, int dir, int avg) { const Picture *pic = (dir == 0) ? &s->last_picture : &s->next_picture; uint8_t *src, *dest; @@ -293,7 +293,7 @@ if (emu) { ff_emulated_edge_mc (s->edge_emu_buffer, src, s->linesize, (width + 1), (height + 1), - mx, my, s->h_edge_pos, s->v_edge_pos); + mx, my, s->h_edge_pos, s->v_edge_pos); src = s->edge_emu_buffer; } if(thirdpel) @@ -302,8 +302,8 @@ (avg ? s->dsp.avg_pixels_tab : s->dsp.put_pixels_tab)[blocksize][dxy](dest, src, s->linesize, height); if (!(s->flags & CODEC_FLAG_GRAY)) { - mx = (mx + (mx < (int) x)) >> 1; - my = (my + (my < (int) y)) >> 1; + mx = (mx + (mx < (int) x)) >> 1; + my = (my + (my < (int) y)) >> 1; width = (width >> 1); height = (height >> 1); blocksize++; @@ -314,7 +314,7 @@ if (emu) { ff_emulated_edge_mc (s->edge_emu_buffer, src, s->uvlinesize, (width + 1), (height + 1), - mx, my, (s->h_edge_pos >> 1), (s->v_edge_pos >> 1)); + mx, my, (s->h_edge_pos >> 1), (s->v_edge_pos >> 1)); src = s->edge_emu_buffer; } if(thirdpel) @@ -344,18 +344,18 @@ k = ((j>>2)&1) + ((i>>1)&2) + ((j>>1)&4) + (i&8); if (mode != PREDICT_MODE) { - pred_motion (h, k, (part_width >> 2), dir, 1, &mx, &my); + pred_motion (h, k, (part_width >> 2), dir, 1, &mx, &my); } else { - mx = s->next_picture.motion_val[0][b_xy][0]<<1; - my = s->next_picture.motion_val[0][b_xy][1]<<1; + mx = s->next_picture.motion_val[0][b_xy][0]<<1; + my = s->next_picture.motion_val[0][b_xy][1]<<1; - if (dir == 0) { - mx = ((mx * h->frame_num_offset) / h->prev_frame_num_offset + 1)>>1; - my = ((my * h->frame_num_offset) / h->prev_frame_num_offset + 1)>>1; - } else { - mx = ((mx * (h->frame_num_offset - h->prev_frame_num_offset)) / h->prev_frame_num_offset + 1)>>1; - my = ((my * (h->frame_num_offset - h->prev_frame_num_offset)) / h->prev_frame_num_offset + 1)>>1; - } + if (dir == 0) { + mx = ((mx * h->frame_num_offset) / h->prev_frame_num_offset + 1)>>1; + my = ((my * h->frame_num_offset) / h->prev_frame_num_offset + 1)>>1; + } else { + mx = ((mx * (h->frame_num_offset - h->prev_frame_num_offset)) / h->prev_frame_num_offset + 1)>>1; + my = ((my * (h->frame_num_offset - h->prev_frame_num_offset)) / h->prev_frame_num_offset + 1)>>1; + } } /* clip motion vector prediction to frame border */ @@ -364,63 +364,63 @@ /* get (optional) motion vector differential */ if (mode == PREDICT_MODE) { - dx = dy = 0; + dx = dy = 0; } else { - dy = svq3_get_se_golomb (&s->gb); - dx = svq3_get_se_golomb (&s->gb); + dy = svq3_get_se_golomb (&s->gb); + dx = svq3_get_se_golomb (&s->gb); - if (dx == INVALID_VLC || dy == INVALID_VLC) { + if (dx == INVALID_VLC || dy == INVALID_VLC) { av_log(h->s.avctx, AV_LOG_ERROR, "invalid MV vlc\n"); - return -1; - } + return -1; + } } /* compute motion vector */ if (mode == THIRDPEL_MODE) { - int fx, fy; - mx = ((mx + 1)>>1) + dx; - my = ((my + 1)>>1) + dy; - fx= ((unsigned)(mx + 0x3000))/3 - 0x1000; - fy= ((unsigned)(my + 0x3000))/3 - 0x1000; - dxy= (mx - 3*fx) + 4*(my - 3*fy); + int fx, fy; + mx = ((mx + 1)>>1) + dx; + my = ((my + 1)>>1) + dy; + fx= ((unsigned)(mx + 0x3000))/3 - 0x1000; + fy= ((unsigned)(my + 0x3000))/3 - 0x1000; + dxy= (mx - 3*fx) + 4*(my - 3*fy); - svq3_mc_dir_part (s, x, y, part_width, part_height, fx, fy, dxy, 1, dir, avg); - mx += mx; - my += my; + svq3_mc_dir_part (s, x, y, part_width, part_height, fx, fy, dxy, 1, dir, avg); + mx += mx; + my += my; } else if (mode == HALFPEL_MODE || mode == PREDICT_MODE) { - mx = ((unsigned)(mx + 1 + 0x3000))/3 + dx - 0x1000; - my = ((unsigned)(my + 1 + 0x3000))/3 + dy - 0x1000; - dxy= (mx&1) + 2*(my&1); + mx = ((unsigned)(mx + 1 + 0x3000))/3 + dx - 0x1000; + my = ((unsigned)(my + 1 + 0x3000))/3 + dy - 0x1000; + dxy= (mx&1) + 2*(my&1); - svq3_mc_dir_part (s, x, y, part_width, part_height, mx>>1, my>>1, dxy, 0, dir, avg); - mx *= 3; - my *= 3; + svq3_mc_dir_part (s, x, y, part_width, part_height, mx>>1, my>>1, dxy, 0, dir, avg); + mx *= 3; + my *= 3; } else { - mx = ((unsigned)(mx + 3 + 0x6000))/6 + dx - 0x1000; - my = ((unsigned)(my + 3 + 0x6000))/6 + dy - 0x1000; + mx = ((unsigned)(mx + 3 + 0x6000))/6 + dx - 0x1000; + my = ((unsigned)(my + 3 + 0x6000))/6 + dy - 0x1000; - svq3_mc_dir_part (s, x, y, part_width, part_height, mx, my, 0, 0, dir, avg); - mx *= 6; - my *= 6; + svq3_mc_dir_part (s, x, y, part_width, part_height, mx, my, 0, 0, dir, avg); + mx *= 6; + my *= 6; } /* update mv_cache */ if (mode != PREDICT_MODE) { - int32_t mv = pack16to32(mx,my); + int32_t mv = pack16to32(mx,my); - if (part_height == 8 && i < 8) { - *(int32_t *) h->mv_cache[dir][scan8[k] + 1*8] = mv; + if (part_height == 8 && i < 8) { + *(int32_t *) h->mv_cache[dir][scan8[k] + 1*8] = mv; - if (part_width == 8 && j < 8) { - *(int32_t *) h->mv_cache[dir][scan8[k] + 1 + 1*8] = mv; - } - } - if (part_width == 8 && j < 8) { - *(int32_t *) h->mv_cache[dir][scan8[k] + 1] = mv; - } - if (part_width == 4 || part_height == 4) { - *(int32_t *) h->mv_cache[dir][scan8[k]] = mv; - } + if (part_width == 8 && j < 8) { + *(int32_t *) h->mv_cache[dir][scan8[k] + 1 + 1*8] = mv; + } + } + if (part_width == 8 && j < 8) { + *(int32_t *) h->mv_cache[dir][scan8[k] + 1] = mv; + } + if (part_width == 4 || part_height == 4) { + *(int32_t *) h->mv_cache[dir][scan8[k]] = mv; + } } /* write back motion vectors */ @@ -440,16 +440,16 @@ const int mb_xy = s->mb_x + s->mb_y*s->mb_stride; const int b_xy = 4*s->mb_x + 4*s->mb_y*h->b_stride; - h->top_samples_available = (s->mb_y == 0) ? 0x33FF : 0xFFFF; - h->left_samples_available = (s->mb_x == 0) ? 0x5F5F : 0xFFFF; - h->topright_samples_available = 0xFFFF; + h->top_samples_available = (s->mb_y == 0) ? 0x33FF : 0xFFFF; + h->left_samples_available = (s->mb_x == 0) ? 0x5F5F : 0xFFFF; + h->topright_samples_available = 0xFFFF; - if (mb_type == 0) { /* SKIP */ + if (mb_type == 0) { /* SKIP */ if (s->pict_type == P_TYPE || s->next_picture.mb_type[mb_xy] == -1) { svq3_mc_dir_part (s, 16*s->mb_x, 16*s->mb_y, 16, 16, 0, 0, 0, 0, 0, 0); if (s->pict_type == B_TYPE) { - svq3_mc_dir_part (s, 16*s->mb_x, 16*s->mb_y, 16, 16, 0, 0, 0, 0, 1, 1); + svq3_mc_dir_part (s, 16*s->mb_x, 16*s->mb_y, 16, 16, 0, 0, 0, 0, 1, 1); } mb_type = MB_TYPE_SKIP; @@ -462,7 +462,7 @@ mb_type = MB_TYPE_16x16; } - } else if (mb_type < 8) { /* INTER */ + } else if (mb_type < 8) { /* INTER */ if (h->thirdpel_flag && h->halfpel_flag == !get_bits (&s->gb, 1)) { mode = THIRDPEL_MODE; } else if (h->halfpel_flag && h->thirdpel_flag == !get_bits (&s->gb, 1)) { @@ -483,107 +483,107 @@ for (m=0; m < 2; m++) { if (s->mb_x > 0 && h->intra4x4_pred_mode[mb_xy - 1][0] != -1) { - for (i=0; i < 4; i++) { - *(uint32_t *) h->mv_cache[m][scan8[0] - 1 + i*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - 1 + i*h->b_stride]; - } + for (i=0; i < 4; i++) { + *(uint32_t *) h->mv_cache[m][scan8[0] - 1 + i*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - 1 + i*h->b_stride]; + } } else { - for (i=0; i < 4; i++) { - *(uint32_t *) h->mv_cache[m][scan8[0] - 1 + i*8] = 0; - } + for (i=0; i < 4; i++) { + *(uint32_t *) h->mv_cache[m][scan8[0] - 1 + i*8] = 0; + } } if (s->mb_y > 0) { - memcpy (h->mv_cache[m][scan8[0] - 1*8], s->current_picture.motion_val[m][b_xy - h->b_stride], 4*2*sizeof(int16_t)); - memset (&h->ref_cache[m][scan8[0] - 1*8], (h->intra4x4_pred_mode[mb_xy - s->mb_stride][4] == -1) ? PART_NOT_AVAILABLE : 1, 4); + memcpy (h->mv_cache[m][scan8[0] - 1*8], s->current_picture.motion_val[m][b_xy - h->b_stride], 4*2*sizeof(int16_t)); + memset (&h->ref_cache[m][scan8[0] - 1*8], (h->intra4x4_pred_mode[mb_xy - s->mb_stride][4] == -1) ? PART_NOT_AVAILABLE : 1, 4); - if (s->mb_x < (s->mb_width - 1)) { - *(uint32_t *) h->mv_cache[m][scan8[0] + 4 - 1*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - h->b_stride + 4]; - h->ref_cache[m][scan8[0] + 4 - 1*8] = - (h->intra4x4_pred_mode[mb_xy - s->mb_stride + 1][0] == -1 || - h->intra4x4_pred_mode[mb_xy - s->mb_stride][4] == -1) ? PART_NOT_AVAILABLE : 1; - }else - h->ref_cache[m][scan8[0] + 4 - 1*8] = PART_NOT_AVAILABLE; - if (s->mb_x > 0) { - *(uint32_t *) h->mv_cache[m][scan8[0] - 1 - 1*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - h->b_stride - 1]; - h->ref_cache[m][scan8[0] - 1 - 1*8] = (h->intra4x4_pred_mode[mb_xy - s->mb_stride - 1][3] == -1) ? PART_NOT_AVAILABLE : 1; - }else - h->ref_cache[m][scan8[0] - 1 - 1*8] = PART_NOT_AVAILABLE; + if (s->mb_x < (s->mb_width - 1)) { + *(uint32_t *) h->mv_cache[m][scan8[0] + 4 - 1*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - h->b_stride + 4]; + h->ref_cache[m][scan8[0] + 4 - 1*8] = + (h->intra4x4_pred_mode[mb_xy - s->mb_stride + 1][0] == -1 || + h->intra4x4_pred_mode[mb_xy - s->mb_stride][4] == -1) ? PART_NOT_AVAILABLE : 1; + }else + h->ref_cache[m][scan8[0] + 4 - 1*8] = PART_NOT_AVAILABLE; + if (s->mb_x > 0) { + *(uint32_t *) h->mv_cache[m][scan8[0] - 1 - 1*8] = *(uint32_t *) s->current_picture.motion_val[m][b_xy - h->b_stride - 1]; + h->ref_cache[m][scan8[0] - 1 - 1*8] = (h->intra4x4_pred_mode[mb_xy - s->mb_stride - 1][3] == -1) ? PART_NOT_AVAILABLE : 1; + }else + h->ref_cache[m][scan8[0] - 1 - 1*8] = PART_NOT_AVAILABLE; }else - memset (&h->ref_cache[m][scan8[0] - 1*8 - 1], PART_NOT_AVAILABLE, 8); + memset (&h->ref_cache[m][scan8[0] - 1*8 - 1], PART_NOT_AVAILABLE, 8); if (s->pict_type != B_TYPE) - break; + break; } /* decode motion vector(s) and form prediction(s) */ if (s->pict_type == P_TYPE) { if(svq3_mc_dir (h, (mb_type - 1), mode, 0, 0) < 0) return -1; - } else { /* B_TYPE */ + } else { /* B_TYPE */ if (mb_type != 2) { - if(svq3_mc_dir (h, 0, mode, 0, 0) < 0) + if(svq3_mc_dir (h, 0, mode, 0, 0) < 0) return -1; } else { - for (i=0; i < 4; i++) { - memset (s->current_picture.motion_val[0][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); - } + for (i=0; i < 4; i++) { + memset (s->current_picture.motion_val[0][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); + } } if (mb_type != 1) { - if(svq3_mc_dir (h, 0, mode, 1, (mb_type == 3)) < 0) + if(svq3_mc_dir (h, 0, mode, 1, (mb_type == 3)) < 0) return -1; } else { - for (i=0; i < 4; i++) { - memset (s->current_picture.motion_val[1][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); - } + for (i=0; i < 4; i++) { + memset (s->current_picture.motion_val[1][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); + } } } mb_type = MB_TYPE_16x16; - } else if (mb_type == 8 || mb_type == 33) { /* INTRA4x4 */ + } else if (mb_type == 8 || mb_type == 33) { /* INTRA4x4 */ memset (h->intra4x4_pred_mode_cache, -1, 8*5*sizeof(int8_t)); if (mb_type == 8) { if (s->mb_x > 0) { - for (i=0; i < 4; i++) { - h->intra4x4_pred_mode_cache[scan8[0] - 1 + i*8] = h->intra4x4_pred_mode[mb_xy - 1][i]; - } - if (h->intra4x4_pred_mode_cache[scan8[0] - 1] == -1) { - h->left_samples_available = 0x5F5F; - } + for (i=0; i < 4; i++) { + h->intra4x4_pred_mode_cache[scan8[0] - 1 + i*8] = h->intra4x4_pred_mode[mb_xy - 1][i]; + } + if (h->intra4x4_pred_mode_cache[scan8[0] - 1] == -1) { + h->left_samples_available = 0x5F5F; + } } if (s->mb_y > 0) { - h->intra4x4_pred_mode_cache[4+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][4]; - h->intra4x4_pred_mode_cache[5+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][5]; - h->intra4x4_pred_mode_cache[6+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][6]; - h->intra4x4_pred_mode_cache[7+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][3]; + h->intra4x4_pred_mode_cache[4+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][4]; + h->intra4x4_pred_mode_cache[5+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][5]; + h->intra4x4_pred_mode_cache[6+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][6]; + h->intra4x4_pred_mode_cache[7+8*0] = h->intra4x4_pred_mode[mb_xy - s->mb_stride][3]; - if (h->intra4x4_pred_mode_cache[4+8*0] == -1) { - h->top_samples_available = 0x33FF; - } + if (h->intra4x4_pred_mode_cache[4+8*0] == -1) { + h->top_samples_available = 0x33FF; + } } /* decode prediction codes for luma blocks */ for (i=0; i < 16; i+=2) { - vlc = svq3_get_ue_golomb (&s->gb); + vlc = svq3_get_ue_golomb (&s->gb); - if (vlc >= 25){ + if (vlc >= 25){ av_log(h->s.avctx, AV_LOG_ERROR, "luma prediction:%d\n", vlc); - return -1; + return -1; } - left = &h->intra4x4_pred_mode_cache[scan8[i] - 1]; - top = &h->intra4x4_pred_mode_cache[scan8[i] - 8]; + left = &h->intra4x4_pred_mode_cache[scan8[i] - 1]; + top = &h->intra4x4_pred_mode_cache[scan8[i] - 8]; - left[1] = svq3_pred_1[top[0] + 1][left[0] + 1][svq3_pred_0[vlc][0]]; - left[2] = svq3_pred_1[top[1] + 1][left[1] + 1][svq3_pred_0[vlc][1]]; + left[1] = svq3_pred_1[top[0] + 1][left[0] + 1][svq3_pred_0[vlc][0]]; + left[2] = svq3_pred_1[top[1] + 1][left[1] + 1][svq3_pred_0[vlc][1]]; - if (left[1] == -1 || left[2] == -1){ + if (left[1] == -1 || left[2] == -1){ av_log(h->s.avctx, AV_LOG_ERROR, "weird prediction\n"); - return -1; + return -1; } } - } else { /* mb_type == 33, DC_128_PRED block type */ + } else { /* mb_type == 33, DC_128_PRED block type */ for (i=0; i < 4; i++) { - memset (&h->intra4x4_pred_mode_cache[scan8[0] + 8*i], DC_PRED, 4); + memset (&h->intra4x4_pred_mode_cache[scan8[0] + 8*i], DC_PRED, 4); } } @@ -604,7 +604,7 @@ } mb_type = MB_TYPE_INTRA4x4; - } else { /* INTRA16x16 */ + } else { /* INTRA16x16 */ dir = i_mb_type_info[mb_type - 8].pred_mode; dir = (dir >> 1) ^ 3*(dir & 1) ^ 1; @@ -623,7 +623,7 @@ } if (s->pict_type == B_TYPE) { for (i=0; i < 4; i++) { - memset (s->current_picture.motion_val[1][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); + memset (s->current_picture.motion_val[1][b_xy + i*h->b_stride], 0, 4*2*sizeof(int16_t)); } } } @@ -664,35 +664,35 @@ for (i=0; i < 4; i++) { if ((cbp & (1 << i))) { - for (j=0; j < 4; j++) { - k = index ? ((j&1) + 2*(i&1) + 2*(j&2) + 4*(i&2)) : (4*i + j); - h->non_zero_count_cache[ scan8[k] ] = 1; + for (j=0; j < 4; j++) { + k = index ? ((j&1) + 2*(i&1) + 2*(j&2) + 4*(i&2)) : (4*i + j); + h->non_zero_count_cache[ scan8[k] ] = 1; - if (svq3_decode_block (&s->gb, &h->mb[16*k], index, type)){ + if (svq3_decode_block (&s->gb, &h->mb[16*k], index, type)){ av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding block\n"); - return -1; + return -1; } - } + } } } if ((cbp & 0x30)) { for (i=0; i < 2; ++i) { - if (svq3_decode_block (&s->gb, &h->mb[16*(16 + 4*i)], 0, 3)){ + if (svq3_decode_block (&s->gb, &h->mb[16*(16 + 4*i)], 0, 3)){ av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding chroma dc block\n"); - return -1; + return -1; } } if ((cbp & 0x20)) { - for (i=0; i < 8; i++) { - h->non_zero_count_cache[ scan8[16+i] ] = 1; + for (i=0; i < 8; i++) { + h->non_zero_count_cache[ scan8[16+i] ] = 1; - if (svq3_decode_block (&s->gb, &h->mb[16*(16 + i)], 1, 1)){ + if (svq3_decode_block (&s->gb, &h->mb[16*(16 + i)], 1, 1)){ av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding chroma ac block\n"); - return -1; + return -1; } - } + } } } } @@ -786,8 +786,8 @@ } static int svq3_decode_frame (AVCodecContext *avctx, - void *data, int *data_size, - uint8_t *buf, int buf_size) { + void *data, int *data_size, + uint8_t *buf, int buf_size) { MpegEncContext *const s = avctx->priv_data; H264Context *const h = avctx->priv_data; int m, mb_type; @@ -833,8 +833,8 @@ /* 'frame size code' and optional 'width, height' */ if (get_bits (&gb, 3) == 7) { - get_bits (&gb, 12); - get_bits (&gb, 12); + get_bits (&gb, 12); + get_bits (&gb, 12); } h->halfpel_flag = get_bits1 (&gb); @@ -852,7 +852,7 @@ get_bits1 (&gb); while (get_bits1 (&gb)) { - get_bits (&gb, 8); + get_bits (&gb, 8); } h->unknown_svq3_flag = get_bits1 (&gb); @@ -935,7 +935,7 @@ for(i=0; i<4; i++){ int j; for(j=-1; j<4; j++) - h->ref_cache[m][scan8[0] + 8*i + j]= 1; + h->ref_cache[m][scan8[0] + 8*i + j]= 1; h->ref_cache[m][scan8[0] + 8*i + j]= PART_NOT_AVAILABLE; } } @@ -944,36 +944,36 @@ for (s->mb_x=0; s->mb_x < s->mb_width; s->mb_x++) { if ( (get_bits_count(&s->gb) + 7) >= s->gb.size_in_bits && - ((get_bits_count(&s->gb) & 7) == 0 || show_bits (&s->gb, (-get_bits_count(&s->gb) & 7)) == 0)) { + ((get_bits_count(&s->gb) & 7) == 0 || show_bits (&s->gb, (-get_bits_count(&s->gb) & 7)) == 0)) { - skip_bits(&s->gb, h->next_slice_index - get_bits_count(&s->gb)); - s->gb.size_in_bits = 8*buf_size; + skip_bits(&s->gb, h->next_slice_index - get_bits_count(&s->gb)); + s->gb.size_in_bits = 8*buf_size; - if (svq3_decode_slice_header (h)) - return -1; + if (svq3_decode_slice_header (h)) + return -1; - /* TODO: support s->mb_skip_run */ + /* TODO: support s->mb_skip_run */ } mb_type = svq3_get_ue_golomb (&s->gb); if (s->pict_type == I_TYPE) { - mb_type += 8; + mb_type += 8; } else if (s->pict_type == B_TYPE && mb_type >= 4) { - mb_type += 4; + mb_type += 4; } if (mb_type > 33 || svq3_decode_mb (h, mb_type)) { - av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding MB %d %d\n", s->mb_x, s->mb_y); - return -1; + av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding MB %d %d\n", s->mb_x, s->mb_y); + return -1; } if (mb_type != 0) { - hl_decode_mb (h); + hl_decode_mb (h); } if (s->pict_type != B_TYPE && !s->low_delay) { - s->current_picture.mb_type[s->mb_x + s->mb_y*s->mb_stride] = - (s->pict_type == P_TYPE && mb_type < 8) ? (mb_type - 1) : -1; + s->current_picture.mb_type[s->mb_x + s->mb_y*s->mb_stride] = + (s->pict_type == P_TYPE && mb_type < 8) ? (mb_type - 1) : -1; } }
--- a/truemotion1.c Wed Dec 21 17:50:40 2005 +0000 +++ b/truemotion1.c Thu Dec 22 01:10:11 2005 +0000 @@ -334,14 +334,14 @@ header.header_size = ((s->buf[0] >> 5) | (s->buf[0] << 3)) & 0x7f; if (s->buf[0] < 0x10) { - av_log(s->avctx, AV_LOG_ERROR, "invalid header size (%d)\n", s->buf[0]); + av_log(s->avctx, AV_LOG_ERROR, "invalid header size (%d)\n", s->buf[0]); return -1; } /* unscramble the header bytes with a XOR operation */ memset(header_buffer, 0, 128); for (i = 1; i < header.header_size; i++) - header_buffer[i - 1] = s->buf[i] ^ s->buf[i + 1]; + header_buffer[i - 1] = s->buf[i] ^ s->buf[i + 1]; header.compression = header_buffer[0]; header.deltaset = header_buffer[1]; @@ -371,7 +371,7 @@ s->flags = FLAG_KEYFRAME; if (s->flags & FLAG_SPRITE) { - av_log(s->avctx, AV_LOG_INFO, "SPRITE frame found, please report the sample to the developers\n"); + av_log(s->avctx, AV_LOG_INFO, "SPRITE frame found, please report the sample to the developers\n"); s->w = header.width; s->h = header.height; s->x = header.xoffset; @@ -381,10 +381,10 @@ s->h = header.ysize; if (header.header_type < 2) { if ((s->w < 213) && (s->h >= 176)) - { + { s->flags |= FLAG_INTERPOLATED; - av_log(s->avctx, AV_LOG_INFO, "INTERPOLATION selected, please report the sample to the developers\n"); - } + av_log(s->avctx, AV_LOG_INFO, "INTERPOLATION selected, please report the sample to the developers\n"); + } } } @@ -412,16 +412,16 @@ if (compression_types[header.compression].algorithm == ALGO_RGB24H) s->avctx->pix_fmt = PIX_FMT_RGBA32; else - s->avctx->pix_fmt = PIX_FMT_RGB555; // RGB565 is supported aswell + s->avctx->pix_fmt = PIX_FMT_RGB555; // RGB565 is supported aswell if ((header.deltaset != s->last_deltaset) || (header.vectable != s->last_vectable)) { if (compression_types[header.compression].algorithm == ALGO_RGB24H) gen_vector_table24(s, sel_vector_table); else - if (s->avctx->pix_fmt == PIX_FMT_RGB555) + if (s->avctx->pix_fmt == PIX_FMT_RGB555) gen_vector_table15(s, sel_vector_table); - else + else gen_vector_table16(s, sel_vector_table); } @@ -445,13 +445,13 @@ s->block_type = compression_types[header.compression].block_type; if (s->avctx->debug & FF_DEBUG_PICT_INFO) - av_log(s->avctx, AV_LOG_INFO, "tables: %d / %d c:%d %dx%d t:%d %s%s%s%s\n", - s->last_deltaset, s->last_vectable, s->compression, s->block_width, - s->block_height, s->block_type, - s->flags & FLAG_KEYFRAME ? " KEY" : "", - s->flags & FLAG_INTERFRAME ? " INTER" : "", - s->flags & FLAG_SPRITE ? " SPRITE" : "", - s->flags & FLAG_INTERPOLATED ? " INTERPOL" : ""); + av_log(s->avctx, AV_LOG_INFO, "tables: %d / %d c:%d %dx%d t:%d %s%s%s%s\n", + s->last_deltaset, s->last_vectable, s->compression, s->block_width, + s->block_height, s->block_type, + s->flags & FLAG_KEYFRAME ? " KEY" : "", + s->flags & FLAG_INTERFRAME ? " INTER" : "", + s->flags & FLAG_SPRITE ? " SPRITE" : "", + s->flags & FLAG_INTERPOLATED ? " INTERPOL" : ""); return header.header_size; } @@ -464,9 +464,9 @@ // FIXME: it may change ? // if (avctx->bits_per_sample == 24) -// avctx->pix_fmt = PIX_FMT_RGB24; +// avctx->pix_fmt = PIX_FMT_RGB24; // else -// avctx->pix_fmt = PIX_FMT_RGB555; +// avctx->pix_fmt = PIX_FMT_RGB555; avctx->has_b_frames = 0; s->frame.data[0] = s->prev_frame.data[0] = NULL;
--- a/tscc.c Wed Dec 21 17:50:40 2005 +0000 +++ b/tscc.c Thu Dec 22 01:10:11 2005 +0000 @@ -110,10 +110,10 @@ for(i = 0; i < p2 * (c->bpp / 8); i++) { *output++ = *src++; } - // RLE8 copy is actually padded - and runs are not! - if(c->bpp == 8 && (p2 & 1)) { - src++; - } + // RLE8 copy is actually padded - and runs are not! + if(c->bpp == 8 && (p2 & 1)) { + src++; + } pos += p2; } else { //Run of pixels int pix[4]; //original pixel
--- a/ulti.c Wed Dec 21 17:50:40 2005 +0000 +++ b/ulti.c Thu Dec 22 01:10:11 2005 +0000 @@ -79,7 +79,7 @@ /* convert Ultimotion YUV block (sixteen 6-bit Y samples and two 4-bit chroma samples) into standard YUV and put it into frame */ static void ulti_convert_yuv(AVFrame *frame, int x, int y, - uint8_t *luma,int chroma) + uint8_t *luma,int chroma) { uint8_t *y_plane, *cr_plane, *cb_plane; int i; @@ -94,31 +94,31 @@ for(i = 0; i < 16; i++){ - y_plane[i & 3] = ulti_lumas[luma[i]]; - if((i & 3) == 3) { //next row - y_plane += frame->linesize[0]; - } + y_plane[i & 3] = ulti_lumas[luma[i]]; + if((i & 3) == 3) { //next row + y_plane += frame->linesize[0]; + } } } /* generate block like in MS Video1 */ static void ulti_pattern(AVFrame *frame, int x, int y, - int f0, int f1, int Y0, int Y1, int chroma) + int f0, int f1, int Y0, int Y1, int chroma) { uint8_t Luma[16]; int mask, i; for(mask = 0x80, i = 0; mask; mask >>= 1, i++) { - if(f0 & mask) - Luma[i] = Y1; - else - Luma[i] = Y0; + if(f0 & mask) + Luma[i] = Y1; + else + Luma[i] = Y0; } for(mask = 0x80, i = 8; mask; mask >>= 1, i++) { - if(f1 & mask) - Luma[i] = Y1; - else - Luma[i] = Y0; + if(f1 & mask) + Luma[i] = Y1; + else + Luma[i] = Y0; } ulti_convert_yuv(frame, x, y, Luma, chroma); @@ -129,70 +129,70 @@ { uint8_t Luma[16]; if(angle & 8) { //reverse order - int t; - angle &= 0x7; - t = Y[0]; - Y[0] = Y[3]; - Y[3] = t; - t = Y[1]; - Y[1] = Y[2]; - Y[2] = t; + int t; + angle &= 0x7; + t = Y[0]; + Y[0] = Y[3]; + Y[3] = t; + t = Y[1]; + Y[1] = Y[2]; + Y[2] = t; } switch(angle){ case 0: - Luma[0] = Y[0]; Luma[1] = Y[1]; Luma[2] = Y[2]; Luma[3] = Y[3]; - Luma[4] = Y[0]; Luma[5] = Y[1]; Luma[6] = Y[2]; Luma[7] = Y[3]; - Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[2]; Luma[11] = Y[3]; - Luma[12] = Y[0]; Luma[13] = Y[1]; Luma[14] = Y[2]; Luma[15] = Y[3]; - break; + Luma[0] = Y[0]; Luma[1] = Y[1]; Luma[2] = Y[2]; Luma[3] = Y[3]; + Luma[4] = Y[0]; Luma[5] = Y[1]; Luma[6] = Y[2]; Luma[7] = Y[3]; + Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[2]; Luma[11] = Y[3]; + Luma[12] = Y[0]; Luma[13] = Y[1]; Luma[14] = Y[2]; Luma[15] = Y[3]; + break; case 1: - Luma[0] = Y[1]; Luma[1] = Y[2]; Luma[2] = Y[3]; Luma[3] = Y[3]; - Luma[4] = Y[0]; Luma[5] = Y[1]; Luma[6] = Y[2]; Luma[7] = Y[3]; - Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[2]; Luma[11] = Y[3]; - Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[1]; Luma[15] = Y[2]; - break; + Luma[0] = Y[1]; Luma[1] = Y[2]; Luma[2] = Y[3]; Luma[3] = Y[3]; + Luma[4] = Y[0]; Luma[5] = Y[1]; Luma[6] = Y[2]; Luma[7] = Y[3]; + Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[2]; Luma[11] = Y[3]; + Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[1]; Luma[15] = Y[2]; + break; case 2: - Luma[0] = Y[1]; Luma[1] = Y[2]; Luma[2] = Y[3]; Luma[3] = Y[3]; - Luma[4] = Y[1]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[3]; - Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[2]; - Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[1]; Luma[15] = Y[2]; - break; + Luma[0] = Y[1]; Luma[1] = Y[2]; Luma[2] = Y[3]; Luma[3] = Y[3]; + Luma[4] = Y[1]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[3]; + Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[2]; + Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[1]; Luma[15] = Y[2]; + break; case 3: - Luma[0] = Y[2]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[3]; - Luma[4] = Y[1]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[3]; - Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[2]; - Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[1]; - break; + Luma[0] = Y[2]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[3]; + Luma[4] = Y[1]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[3]; + Luma[8] = Y[0]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[2]; + Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[1]; + break; case 4: - Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[3]; - Luma[4] = Y[2]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[2]; - Luma[8] = Y[1]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[1]; - Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[0]; - break; + Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[3]; + Luma[4] = Y[2]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[2]; + Luma[8] = Y[1]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[1]; + Luma[12] = Y[0]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[0]; + break; case 5: - Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[2]; - Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[1]; - Luma[8] = Y[2]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[0]; - Luma[12] = Y[1]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[0]; - break; + Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[3]; Luma[3] = Y[2]; + Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[2]; Luma[7] = Y[1]; + Luma[8] = Y[2]; Luma[9] = Y[1]; Luma[10] = Y[1]; Luma[11] = Y[0]; + Luma[12] = Y[1]; Luma[13] = Y[0]; Luma[14] = Y[0]; Luma[15] = Y[0]; + break; case 6: - Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[2]; Luma[3] = Y[2]; - Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[1]; Luma[7] = Y[1]; - Luma[8] = Y[2]; Luma[9] = Y[2]; Luma[10] = Y[1]; Luma[11] = Y[0]; - Luma[12] = Y[1]; Luma[13] = Y[1]; Luma[14] = Y[0]; Luma[15] = Y[0]; - break; + Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[2]; Luma[3] = Y[2]; + Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[1]; Luma[7] = Y[1]; + Luma[8] = Y[2]; Luma[9] = Y[2]; Luma[10] = Y[1]; Luma[11] = Y[0]; + Luma[12] = Y[1]; Luma[13] = Y[1]; Luma[14] = Y[0]; Luma[15] = Y[0]; + break; case 7: - Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[2]; Luma[3] = Y[1]; - Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[1]; Luma[7] = Y[0]; - Luma[8] = Y[3]; Luma[9] = Y[2]; Luma[10] = Y[1]; Luma[11] = Y[0]; - Luma[12] = Y[2]; Luma[13] = Y[1]; Luma[14] = Y[0]; Luma[15] = Y[0]; - break; + Luma[0] = Y[3]; Luma[1] = Y[3]; Luma[2] = Y[2]; Luma[3] = Y[1]; + Luma[4] = Y[3]; Luma[5] = Y[2]; Luma[6] = Y[1]; Luma[7] = Y[0]; + Luma[8] = Y[3]; Luma[9] = Y[2]; Luma[10] = Y[1]; Luma[11] = Y[0]; + Luma[12] = Y[2]; Luma[13] = Y[1]; Luma[14] = Y[0]; Luma[15] = Y[0]; + break; default: - Luma[0] = Y[0]; Luma[1] = Y[0]; Luma[2] = Y[1]; Luma[3] = Y[1]; - Luma[4] = Y[0]; Luma[5] = Y[0]; Luma[6] = Y[1]; Luma[7] = Y[1]; - Luma[8] = Y[2]; Luma[9] = Y[2]; Luma[10] = Y[3]; Luma[11] = Y[3]; - Luma[12] = Y[2]; Luma[13] = Y[2]; Luma[14] = Y[3]; Luma[15] = Y[3]; - break; + Luma[0] = Y[0]; Luma[1] = Y[0]; Luma[2] = Y[1]; Luma[3] = Y[1]; + Luma[4] = Y[0]; Luma[5] = Y[0]; Luma[6] = Y[1]; Luma[7] = Y[1]; + Luma[8] = Y[2]; Luma[9] = Y[2]; Luma[10] = Y[3]; Luma[11] = Y[3]; + Luma[12] = Y[2]; Luma[13] = Y[2]; Luma[14] = Y[3]; Luma[15] = Y[3]; + break; } ulti_convert_yuv(frame, x, y, Luma, chroma); @@ -224,178 +224,178 @@ } while(!done) { - int idx; - if(blocks >= s->blocks || y >= s->height) - break;//all blocks decoded + int idx; + if(blocks >= s->blocks || y >= s->height) + break;//all blocks decoded - idx = *buf++; - if((idx & 0xF8) == 0x70) { - switch(idx) { - case 0x70: //change modifier - modifier = *buf++; - if(modifier>1) - av_log(avctx, AV_LOG_INFO, "warning: modifier must be 0 or 1, got %i\n", modifier); - break; - case 0x71: // set uniq flag - uniq = 1; - break; - case 0x72: //toggle mode - mode = !mode; - break; - case 0x73: //end-of-frame - done = 1; - break; - case 0x74: //skip some blocks - skip = *buf++; - if ((blocks + skip) >= s->blocks) - break; - blocks += skip; - x += skip * 8; - while(x >= s->width) { - x -= s->width; - y += 8; - } - break; - default: - av_log(avctx, AV_LOG_INFO, "warning: unknown escape 0x%02X\n", idx); - } - } else { //handle one block - int code; - int cf; - int angle = 0; - uint8_t Y[4]; // luma samples of block - int tx = 0, ty = 0; //coords of subblock - int chroma = 0; - if (mode || uniq) { - uniq = 0; - cf = 1; - chroma = 0; - } else { - cf = 0; - if (idx) - chroma = *buf++; - } - for (i = 0; i < 4; i++) { // for every subblock - code = (idx >> (6 - i*2)) & 3; //extract 2 bits - if(!code) //skip subblock - continue; - if(cf) - chroma = *buf++; - tx = x + block_coords[i * 2]; - ty = y + block_coords[(i * 2) + 1]; - switch(code) { - case 1: - tmp = *buf++; + idx = *buf++; + if((idx & 0xF8) == 0x70) { + switch(idx) { + case 0x70: //change modifier + modifier = *buf++; + if(modifier>1) + av_log(avctx, AV_LOG_INFO, "warning: modifier must be 0 or 1, got %i\n", modifier); + break; + case 0x71: // set uniq flag + uniq = 1; + break; + case 0x72: //toggle mode + mode = !mode; + break; + case 0x73: //end-of-frame + done = 1; + break; + case 0x74: //skip some blocks + skip = *buf++; + if ((blocks + skip) >= s->blocks) + break; + blocks += skip; + x += skip * 8; + while(x >= s->width) { + x -= s->width; + y += 8; + } + break; + default: + av_log(avctx, AV_LOG_INFO, "warning: unknown escape 0x%02X\n", idx); + } + } else { //handle one block + int code; + int cf; + int angle = 0; + uint8_t Y[4]; // luma samples of block + int tx = 0, ty = 0; //coords of subblock + int chroma = 0; + if (mode || uniq) { + uniq = 0; + cf = 1; + chroma = 0; + } else { + cf = 0; + if (idx) + chroma = *buf++; + } + for (i = 0; i < 4; i++) { // for every subblock + code = (idx >> (6 - i*2)) & 3; //extract 2 bits + if(!code) //skip subblock + continue; + if(cf) + chroma = *buf++; + tx = x + block_coords[i * 2]; + ty = y + block_coords[(i * 2) + 1]; + switch(code) { + case 1: + tmp = *buf++; - angle = angle_by_index[(tmp >> 6) & 0x3]; + angle = angle_by_index[(tmp >> 6) & 0x3]; - Y[0] = tmp & 0x3F; - Y[1] = Y[0]; + Y[0] = tmp & 0x3F; + Y[1] = Y[0]; - if (angle) { - Y[2] = Y[0]+1; - if (Y[2] > 0x3F) - Y[2] = 0x3F; - Y[3] = Y[2]; - } else { - Y[2] = Y[0]; - Y[3] = Y[0]; - } - break; + if (angle) { + Y[2] = Y[0]+1; + if (Y[2] > 0x3F) + Y[2] = 0x3F; + Y[3] = Y[2]; + } else { + Y[2] = Y[0]; + Y[3] = Y[0]; + } + break; - case 2: - if (modifier) { // unpack four luma samples - tmp = (*buf++) << 16; - tmp += (*buf++) << 8; - tmp += *buf++; + case 2: + if (modifier) { // unpack four luma samples + tmp = (*buf++) << 16; + tmp += (*buf++) << 8; + tmp += *buf++; - Y[0] = (tmp >> 18) & 0x3F; - Y[1] = (tmp >> 12) & 0x3F; - Y[2] = (tmp >> 6) & 0x3F; - Y[3] = tmp & 0x3F; - angle = 16; - } else { // retrieve luma samples from codebook - tmp = (*buf++) << 8; - tmp += (*buf++); + Y[0] = (tmp >> 18) & 0x3F; + Y[1] = (tmp >> 12) & 0x3F; + Y[2] = (tmp >> 6) & 0x3F; + Y[3] = tmp & 0x3F; + angle = 16; + } else { // retrieve luma samples from codebook + tmp = (*buf++) << 8; + tmp += (*buf++); - angle = (tmp >> 12) & 0xF; - tmp &= 0xFFF; - tmp <<= 2; - Y[0] = s->ulti_codebook[tmp]; - Y[1] = s->ulti_codebook[tmp + 1]; - Y[2] = s->ulti_codebook[tmp + 2]; - Y[3] = s->ulti_codebook[tmp + 3]; - } - break; + angle = (tmp >> 12) & 0xF; + tmp &= 0xFFF; + tmp <<= 2; + Y[0] = s->ulti_codebook[tmp]; + Y[1] = s->ulti_codebook[tmp + 1]; + Y[2] = s->ulti_codebook[tmp + 2]; + Y[3] = s->ulti_codebook[tmp + 3]; + } + break; - case 3: - if (modifier) { // all 16 luma samples - uint8_t Luma[16]; + case 3: + if (modifier) { // all 16 luma samples + uint8_t Luma[16]; - tmp = (*buf++) << 16; - tmp += (*buf++) << 8; - tmp += *buf++; - Luma[0] = (tmp >> 18) & 0x3F; - Luma[1] = (tmp >> 12) & 0x3F; - Luma[2] = (tmp >> 6) & 0x3F; - Luma[3] = tmp & 0x3F; + tmp = (*buf++) << 16; + tmp += (*buf++) << 8; + tmp += *buf++; + Luma[0] = (tmp >> 18) & 0x3F; + Luma[1] = (tmp >> 12) & 0x3F; + Luma[2] = (tmp >> 6) & 0x3F; + Luma[3] = tmp & 0x3F; - tmp = (*buf++) << 16; - tmp += (*buf++) << 8; - tmp += *buf++; - Luma[4] = (tmp >> 18) & 0x3F; - Luma[5] = (tmp >> 12) & 0x3F; - Luma[6] = (tmp >> 6) & 0x3F; - Luma[7] = tmp & 0x3F; + tmp = (*buf++) << 16; + tmp += (*buf++) << 8; + tmp += *buf++; + Luma[4] = (tmp >> 18) & 0x3F; + Luma[5] = (tmp >> 12) & 0x3F; + Luma[6] = (tmp >> 6) & 0x3F; + Luma[7] = tmp & 0x3F; - tmp = (*buf++) << 16; - tmp += (*buf++) << 8; - tmp += *buf++; - Luma[8] = (tmp >> 18) & 0x3F; - Luma[9] = (tmp >> 12) & 0x3F; - Luma[10] = (tmp >> 6) & 0x3F; - Luma[11] = tmp & 0x3F; + tmp = (*buf++) << 16; + tmp += (*buf++) << 8; + tmp += *buf++; + Luma[8] = (tmp >> 18) & 0x3F; + Luma[9] = (tmp >> 12) & 0x3F; + Luma[10] = (tmp >> 6) & 0x3F; + Luma[11] = tmp & 0x3F; - tmp = (*buf++) << 16; - tmp += (*buf++) << 8; - tmp += *buf++; - Luma[12] = (tmp >> 18) & 0x3F; - Luma[13] = (tmp >> 12) & 0x3F; - Luma[14] = (tmp >> 6) & 0x3F; - Luma[15] = tmp & 0x3F; + tmp = (*buf++) << 16; + tmp += (*buf++) << 8; + tmp += *buf++; + Luma[12] = (tmp >> 18) & 0x3F; + Luma[13] = (tmp >> 12) & 0x3F; + Luma[14] = (tmp >> 6) & 0x3F; + Luma[15] = tmp & 0x3F; - ulti_convert_yuv(&s->frame, tx, ty, Luma, chroma); - } else { - tmp = *buf++; - if(tmp & 0x80) { - angle = (tmp >> 4) & 0x7; - tmp = (tmp << 8) + *buf++; - Y[0] = (tmp >> 6) & 0x3F; - Y[1] = tmp & 0x3F; - Y[2] = (*buf++) & 0x3F; - Y[3] = (*buf++) & 0x3F; - ulti_grad(&s->frame, tx, ty, Y, chroma, angle); //draw block - } else { // some patterns - int f0, f1; - f0 = *buf++; - f1 = tmp; - Y[0] = (*buf++) & 0x3F; - Y[1] = (*buf++) & 0x3F; - ulti_pattern(&s->frame, tx, ty, f1, f0, Y[0], Y[1], chroma); - } - } - break; - } - if(code != 3) - ulti_grad(&s->frame, tx, ty, Y, chroma, angle); // draw block - } - blocks++; - x += 8; - if(x >= s->width) { - x = 0; - y += 8; - } - } + ulti_convert_yuv(&s->frame, tx, ty, Luma, chroma); + } else { + tmp = *buf++; + if(tmp & 0x80) { + angle = (tmp >> 4) & 0x7; + tmp = (tmp << 8) + *buf++; + Y[0] = (tmp >> 6) & 0x3F; + Y[1] = tmp & 0x3F; + Y[2] = (*buf++) & 0x3F; + Y[3] = (*buf++) & 0x3F; + ulti_grad(&s->frame, tx, ty, Y, chroma, angle); //draw block + } else { // some patterns + int f0, f1; + f0 = *buf++; + f1 = tmp; + Y[0] = (*buf++) & 0x3F; + Y[1] = (*buf++) & 0x3F; + ulti_pattern(&s->frame, tx, ty, f1, f0, Y[0], Y[1], chroma); + } + } + break; + } + if(code != 3) + ulti_grad(&s->frame, tx, ty, Y, chroma, angle); // draw block + } + blocks++; + x += 8; + if(x >= s->width) { + x = 0; + y += 8; + } + } } *data_size=sizeof(AVFrame);
--- a/utils.c Wed Dec 21 17:50:40 2005 +0000 +++ b/utils.c Thu Dec 22 01:10:11 2005 +0000 @@ -1188,7 +1188,7 @@ static int inited = 0; if (inited != 0) - return; + return; inited = 1; dsputil_static_init(); @@ -1241,10 +1241,10 @@ static int print_prefix=1; AVClass* avc= ptr ? *(AVClass**)ptr : NULL; if(level>av_log_level) - return; + return; #undef fprintf if(print_prefix && avc) { - fprintf(stderr, "[%s @ %p]", avc->item_name(ptr), avc); + fprintf(stderr, "[%s @ %p]", avc->item_name(ptr), avc); } #define fprintf please_use_av_log
--- a/vmdav.c Wed Dec 21 17:50:40 2005 +0000 +++ b/vmdav.c Thu Dec 22 01:10:11 2005 +0000 @@ -431,7 +431,7 @@ s->block_align = avctx->block_align; av_log(s->avctx, AV_LOG_DEBUG, "%d channels, %d bits/sample, block align = %d, sample rate = %d\n", - s->channels, s->bits, s->block_align, avctx->sample_rate); + s->channels, s->bits, s->block_align, avctx->sample_rate); /* set up the steps8 and steps16 tables */ for (i = 0; i < 8; i++) { @@ -489,7 +489,7 @@ int i; if (silence) - av_log(s->avctx, AV_LOG_INFO, "silent block!\n"); + av_log(s->avctx, AV_LOG_INFO, "silent block!\n"); if (s->channels == 2) { /* stereo handling */
--- a/vp3.c Wed Dec 21 17:50:40 2005 +0000 +++ b/vp3.c Thu Dec 22 01:10:11 2005 +0000 @@ -2217,9 +2217,9 @@ int c_superblock_count; if (avctx->codec_tag == MKTAG('V','P','3','0')) - s->version = 0; + s->version = 0; else - s->version = 1; + s->version = 1; s->avctx = avctx; s->width = (avctx->width + 15) & 0xFFFFFFF0; @@ -2288,18 +2288,18 @@ if (!s->theora_tables) { - for (i = 0; i < 64; i++) - s->coded_dc_scale_factor[i] = vp31_dc_scale_factor[i]; - for (i = 0; i < 64; i++) - s->coded_ac_scale_factor[i] = vp31_ac_scale_factor[i]; - for (i = 0; i < 64; i++) - s->coded_intra_y_dequant[i] = vp31_intra_y_dequant[i]; - for (i = 0; i < 64; i++) - s->coded_intra_c_dequant[i] = vp31_intra_c_dequant[i]; - for (i = 0; i < 64; i++) - s->coded_inter_dequant[i] = vp31_inter_dequant[i]; - for (i = 0; i < 64; i++) - s->filter_limit_values[i] = vp31_filter_limit_values[i]; + for (i = 0; i < 64; i++) + s->coded_dc_scale_factor[i] = vp31_dc_scale_factor[i]; + for (i = 0; i < 64; i++) + s->coded_ac_scale_factor[i] = vp31_ac_scale_factor[i]; + for (i = 0; i < 64; i++) + s->coded_intra_y_dequant[i] = vp31_intra_y_dequant[i]; + for (i = 0; i < 64; i++) + s->coded_intra_c_dequant[i] = vp31_intra_c_dequant[i]; + for (i = 0; i < 64; i++) + s->coded_inter_dequant[i] = vp31_inter_dequant[i]; + for (i = 0; i < 64; i++) + s->filter_limit_values[i] = vp31_filter_limit_values[i]; /* init VLC tables */ for (i = 0; i < 16; i++) { @@ -2408,40 +2408,40 @@ if (s->theora && get_bits1(&gb)) { #if 1 - av_log(avctx, AV_LOG_ERROR, "Header packet passed to frame decoder, skipping\n"); - return -1; + av_log(avctx, AV_LOG_ERROR, "Header packet passed to frame decoder, skipping\n"); + return -1; #else - int ptype = get_bits(&gb, 7); - - skip_bits(&gb, 6*8); /* "theora" */ - - switch(ptype) - { - case 1: - theora_decode_comments(avctx, gb); - break; - case 2: - theora_decode_tables(avctx, gb); - init_dequantizer(s); - break; - default: - av_log(avctx, AV_LOG_ERROR, "Unknown Theora config packet: %d\n", ptype); - } - return buf_size; + int ptype = get_bits(&gb, 7); + + skip_bits(&gb, 6*8); /* "theora" */ + + switch(ptype) + { + case 1: + theora_decode_comments(avctx, gb); + break; + case 2: + theora_decode_tables(avctx, gb); + init_dequantizer(s); + break; + default: + av_log(avctx, AV_LOG_ERROR, "Unknown Theora config packet: %d\n", ptype); + } + return buf_size; #endif } s->keyframe = !get_bits1(&gb); if (!s->theora) - skip_bits(&gb, 1); + skip_bits(&gb, 1); s->last_quality_index = s->quality_index; s->quality_index = get_bits(&gb, 6); if (s->theora >= 0x030200) skip_bits1(&gb); if (s->avctx->debug & FF_DEBUG_PICT_INFO) - av_log(s->avctx, AV_LOG_INFO, " VP3 %sframe #%d: Q index = %d\n", - s->keyframe?"key":"", counter, s->quality_index); + av_log(s->avctx, AV_LOG_INFO, " VP3 %sframe #%d: Q index = %d\n", + s->keyframe?"key":"", counter, s->quality_index); counter++; if (s->quality_index != s->last_quality_index) { @@ -2450,23 +2450,23 @@ } if (s->keyframe) { - if (!s->theora) - { - skip_bits(&gb, 4); /* width code */ - skip_bits(&gb, 4); /* height code */ - if (s->version) - { - s->version = get_bits(&gb, 5); - if (counter == 1) - av_log(s->avctx, AV_LOG_DEBUG, "VP version: %d\n", s->version); - } - } - if (s->version || s->theora) - { - if (get_bits1(&gb)) - av_log(s->avctx, AV_LOG_ERROR, "Warning, unsupported keyframe coding type?!\n"); - skip_bits(&gb, 2); /* reserved? */ - } + if (!s->theora) + { + skip_bits(&gb, 4); /* width code */ + skip_bits(&gb, 4); /* height code */ + if (s->version) + { + s->version = get_bits(&gb, 5); + if (counter == 1) + av_log(s->avctx, AV_LOG_DEBUG, "VP version: %d\n", s->version); + } + } + if (s->version || s->theora) + { + if (get_bits1(&gb)) + av_log(s->avctx, AV_LOG_ERROR, "Warning, unsupported keyframe coding type?!\n"); + skip_bits(&gb, 2); /* reserved? */ + } if (s->last_frame.data[0] == s->golden_frame.data[0]) { if (s->golden_frame.data[0]) @@ -2490,12 +2490,12 @@ /* time to figure out pixel addresses? */ if (!s->pixel_addresses_inited) - { - if (!s->flipped_image) - vp3_calculate_pixel_addresses(s); - else - theora_calculate_pixel_addresses(s); - } + { + if (!s->flipped_image) + vp3_calculate_pixel_addresses(s); + else + theora_calculate_pixel_addresses(s); + } } else { /* allocate a new current frame */ s->current_frame.reference = 3; @@ -2655,7 +2655,7 @@ minor = get_bits(&gb, 8); /* version minor */ micro = get_bits(&gb, 8); /* version micro */ av_log(avctx, AV_LOG_INFO, "Theora bitstream version %d.%d.%d\n", - major, minor, micro); + major, minor, micro); /* FIXME: endianess? */ s->theora = (major << 16) | (minor << 8) | micro; @@ -2664,7 +2664,7 @@ /* but previous versions have the image flipped relative to vp3 */ if (s->theora < 0x030200) { - s->flipped_image = 1; + s->flipped_image = 1; av_log(avctx, AV_LOG_DEBUG, "Old (<alpha3) Theora bitstream, flipped image\n"); } @@ -2679,19 +2679,19 @@ if (s->theora >= 0x030400) { - skip_bits(&gb, 32); /* total number of superblocks in a frame */ - // fixme, the next field is 36bits long - skip_bits(&gb, 32); /* total number of blocks in a frame */ - skip_bits(&gb, 4); /* total number of blocks in a frame */ - skip_bits(&gb, 32); /* total number of macroblocks in a frame */ - - skip_bits(&gb, 24); /* frame width */ - skip_bits(&gb, 24); /* frame height */ + skip_bits(&gb, 32); /* total number of superblocks in a frame */ + // fixme, the next field is 36bits long + skip_bits(&gb, 32); /* total number of blocks in a frame */ + skip_bits(&gb, 4); /* total number of blocks in a frame */ + skip_bits(&gb, 32); /* total number of macroblocks in a frame */ + + skip_bits(&gb, 24); /* frame width */ + skip_bits(&gb, 24); /* frame height */ } else { - skip_bits(&gb, 24); /* frame width */ - skip_bits(&gb, 24); /* frame height */ + skip_bits(&gb, 24); /* frame width */ + skip_bits(&gb, 24); /* frame height */ } skip_bits(&gb, 8); /* offset x */ @@ -2703,20 +2703,20 @@ skip_bits(&gb, 24); /* aspect denumerator */ if (s->theora < 0x030200) - skip_bits(&gb, 5); /* keyframe frequency force */ + skip_bits(&gb, 5); /* keyframe frequency force */ skip_bits(&gb, 8); /* colorspace */ if (s->theora >= 0x030400) - skip_bits(&gb, 2); /* pixel format: 420,res,422,444 */ + skip_bits(&gb, 2); /* pixel format: 420,res,422,444 */ skip_bits(&gb, 24); /* bitrate */ skip_bits(&gb, 6); /* quality hint */ if (s->theora >= 0x030200) { - skip_bits(&gb, 5); /* keyframe frequency force */ - - if (s->theora < 0x030400) - skip_bits(&gb, 5); /* spare bits */ + skip_bits(&gb, 5); /* keyframe frequency force */ + + if (s->theora < 0x030400) + skip_bits(&gb, 5); /* spare bits */ } // align_get_bits(&gb); @@ -2744,35 +2744,35 @@ if (s->theora <= 0x030200) { - int i, comments; - - // vendor string - len = get_bits_long(&gb, 32); - len = le2me_32(len); - while(len--) - skip_bits(&gb, 8); - - // user comments - comments = get_bits_long(&gb, 32); - comments = le2me_32(comments); - for (i = 0; i < comments; i++) - { - len = get_bits_long(&gb, 32); - len = be2me_32(len); - while(len--) - skip_bits(&gb, 8); - } + int i, comments; + + // vendor string + len = get_bits_long(&gb, 32); + len = le2me_32(len); + while(len--) + skip_bits(&gb, 8); + + // user comments + comments = get_bits_long(&gb, 32); + comments = le2me_32(comments); + for (i = 0; i < comments; i++) + { + len = get_bits_long(&gb, 32); + len = be2me_32(len); + while(len--) + skip_bits(&gb, 8); + } } else { - do { - len = get_bits_long(&gb, 32); - len = le2me_32(len); - if (len <= 0) - break; - while (len--) - skip_bits(&gb, 8); - } while (1); + do { + len = get_bits_long(&gb, 32); + len = le2me_32(len); + if (len <= 0) + break; + while (len--) + skip_bits(&gb, 8); + } while (1); } return 0; } @@ -2795,7 +2795,7 @@ n = 16; /* quality threshold table */ for (i = 0; i < 64; i++) - s->coded_ac_scale_factor[i] = get_bits(&gb, n); + s->coded_ac_scale_factor[i] = get_bits(&gb, n); if (s->theora >= 0x030200) n = get_bits(&gb, 4) + 1; @@ -2803,7 +2803,7 @@ n = 16; /* dc scale factor table */ for (i = 0; i < 64; i++) - s->coded_dc_scale_factor[i] = get_bits(&gb, n); + s->coded_dc_scale_factor[i] = get_bits(&gb, n); if (s->theora >= 0x030200) matrices = get_bits(&gb, 9) + 1; @@ -2815,21 +2815,21 @@ } /* y coeffs */ for (i = 0; i < 64; i++) - s->coded_intra_y_dequant[i] = get_bits(&gb, 8); + s->coded_intra_y_dequant[i] = get_bits(&gb, 8); /* uv coeffs */ for (i = 0; i < 64; i++) - s->coded_intra_c_dequant[i] = get_bits(&gb, 8); + s->coded_intra_c_dequant[i] = get_bits(&gb, 8); /* inter coeffs */ for (i = 0; i < 64; i++) - s->coded_inter_dequant[i] = get_bits(&gb, 8); + s->coded_inter_dequant[i] = get_bits(&gb, 8); /* skip unknown matrices */ n = matrices - 3; while(n--) - for (i = 0; i < 64; i++) - skip_bits(&gb, 8); + for (i = 0; i < 64; i++) + skip_bits(&gb, 8); for (i = 0; i <= 1; i++) { for (n = 0; n <= 2; n++) { @@ -2851,8 +2851,8 @@ } if (qi > 63) { av_log(avctx, AV_LOG_ERROR, "invalid qi %d > 63\n", qi); - return -1; - } + return -1; + } } } } @@ -2887,7 +2887,7 @@ if (!avctx->extradata_size) { av_log(avctx, AV_LOG_ERROR, "Missing extradata!\n"); - return -1; + return -1; } for(i=0;i<3;i++) { @@ -2903,7 +2903,7 @@ if (!(ptype & 0x80)) { av_log(avctx, AV_LOG_ERROR, "Invalid extradata!\n"); - return -1; + return -1; } // FIXME: check for this aswell @@ -2913,17 +2913,17 @@ { case 0x80: theora_decode_header(avctx, gb); - break; - case 0x81: + break; + case 0x81: // FIXME: is this needed? it breaks sometimes -// theora_decode_comments(avctx, gb); - break; - case 0x82: - theora_decode_tables(avctx, gb); - break; - default: - av_log(avctx, AV_LOG_ERROR, "Unknown Theora config packet: %d\n", ptype&~0x80); - break; +// theora_decode_comments(avctx, gb); + break; + case 0x82: + theora_decode_tables(avctx, gb); + break; + default: + av_log(avctx, AV_LOG_ERROR, "Unknown Theora config packet: %d\n", ptype&~0x80); + break; } }
--- a/wmv2.c Wed Dec 21 17:50:40 2005 +0000 +++ b/wmv2.c Thu Dec 22 01:10:11 2005 +0000 @@ -206,13 +206,13 @@ handle_slices(s); if (!s->mb_intra) { - /* compute cbp */ + /* compute cbp */ set_stat(ST_INTER_MB); - cbp = 0; - for (i = 0; i < 6; i++) { - if (s->block_last_index[i] >= 0) - cbp |= 1 << (5 - i); - } + cbp = 0; + for (i = 0; i < 6; i++) { + if (s->block_last_index[i] >= 0) + cbp |= 1 << (5 - i); + } put_bits(&s->pb, wmv2_inter_table[w->cbp_table_index][cbp + 64][1], @@ -223,10 +223,10 @@ msmpeg4_encode_motion(s, motion_x - pred_x, motion_y - pred_y); } else { - /* compute cbp */ - cbp = 0; + /* compute cbp */ + cbp = 0; coded_cbp = 0; - for (i = 0; i < 6; i++) { + for (i = 0; i < 6; i++) { int val, pred; val = (s->block_last_index[i] >= 1); cbp |= val << (5 - i); @@ -237,7 +237,7 @@ val = val ^ pred; } coded_cbp |= val << (5 - i); - } + } #if 0 if (coded_cbp) printf("cbp=%x %x\n", cbp, coded_cbp); @@ -253,7 +253,7 @@ wmv2_inter_table[w->cbp_table_index][cbp][0]); } set_stat(ST_INTRA_MB); - put_bits(&s->pb, 1, 0); /* no AC prediction yet */ + put_bits(&s->pb, 1, 0); /* no AC prediction yet */ if(s->inter_intra_pred){ s->h263_aic_dir=0; put_bits(&s->pb, table_inter_intra[s->h263_aic_dir][1], table_inter_intra[s->h263_aic_dir][0]); @@ -400,11 +400,11 @@ s->inter_intra_pred= 0; s->no_rounding = 1; if(s->avctx->debug&FF_DEBUG_PICT_INFO){ - av_log(s->avctx, AV_LOG_DEBUG, "qscale:%d rlc:%d rl:%d dc:%d mbrl:%d j_type:%d \n", - s->qscale, - s->rl_chroma_table_index, - s->rl_table_index, - s->dc_table_index, + av_log(s->avctx, AV_LOG_DEBUG, "qscale:%d rlc:%d rl:%d dc:%d mbrl:%d j_type:%d \n", + s->qscale, + s->rl_chroma_table_index, + s->rl_table_index, + s->dc_table_index, s->per_mb_rl_table, w->j_type); } @@ -451,10 +451,10 @@ if(s->avctx->debug&FF_DEBUG_PICT_INFO){ av_log(s->avctx, AV_LOG_DEBUG, "rl:%d rlc:%d dc:%d mv:%d mbrl:%d qp:%d mspel:%d per_mb_abt:%d abt_type:%d cbp:%d ii:%d\n", - s->rl_table_index, - s->rl_chroma_table_index, - s->dc_table_index, - s->mv_table_index, + s->rl_table_index, + s->rl_chroma_table_index, + s->dc_table_index, + s->mv_table_index, s->per_mb_rl_table, s->qscale, s->mspel, @@ -734,7 +734,7 @@ code = get_vlc2(&s->gb, mb_non_intra_vlc[w->cbp_table_index].table, MB_NON_INTRA_VLC_BITS, 3); if (code < 0) return -1; - s->mb_intra = (~code & 0x40) >> 6; + s->mb_intra = (~code & 0x40) >> 6; cbp = code & 0x3f; } else { @@ -787,10 +787,10 @@ for (i = 0; i < 6; i++) { if (wmv2_decode_inter_block(w, block[i], i, (cbp >> (5 - i)) & 1) < 0) - { - av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding inter block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); - return -1; - } + { + av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding inter block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); + return -1; + } } } else { //if(s->pict_type==P_TYPE) @@ -809,10 +809,10 @@ s->dsp.clear_blocks(s->block[0]); for (i = 0; i < 6; i++) { if (msmpeg4_decode_block(s, block[i], i, (cbp >> (5 - i)) & 1, NULL) < 0) - { - av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding intra block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); - return -1; - } + { + av_log(s->avctx, AV_LOG_ERROR, "\nerror while decoding intra block: %d x %d (%d)\n", s->mb_x, s->mb_y, i); + return -1; + } } }
--- a/x264.c Wed Dec 21 17:50:40 2005 +0000 +++ b/x264.c Thu Dec 22 01:10:11 2005 +0000 @@ -32,14 +32,14 @@ X264_log(void *p, int level, const char *fmt, va_list args) { static const int level_map[] = { - [X264_LOG_ERROR] = AV_LOG_ERROR, - [X264_LOG_WARNING] = AV_LOG_ERROR, - [X264_LOG_INFO] = AV_LOG_INFO, - [X264_LOG_DEBUG] = AV_LOG_DEBUG + [X264_LOG_ERROR] = AV_LOG_ERROR, + [X264_LOG_WARNING] = AV_LOG_ERROR, + [X264_LOG_INFO] = AV_LOG_INFO, + [X264_LOG_DEBUG] = AV_LOG_DEBUG }; if(level < 0 || level > X264_LOG_DEBUG) - return; + return; av_vlog(p, level_map[level], fmt, args); } @@ -52,10 +52,10 @@ int i; for(i = 0; i < nnal; i++){ - int s = x264_nal_encode(p, &size, 1, nals + i); - if(s < 0) - return -1; - p += s; + int s = x264_nal_encode(p, &size, 1, nals + i); + if(s < 0) + return -1; + p += s; } return p - buf; @@ -74,19 +74,19 @@ x4->pic.img.i_plane = 3; for(i = 0; i < 3; i++){ - x4->pic.img.plane[i] = frame->data[i]; - x4->pic.img.i_stride[i] = frame->linesize[i]; + x4->pic.img.plane[i] = frame->data[i]; + x4->pic.img.i_stride[i] = frame->linesize[i]; } x4->pic.i_pts = frame->pts; x4->pic.i_type = X264_TYPE_AUTO; if(x264_encoder_encode(x4->enc, &nal, &nnal, &x4->pic, &pic_out)) - return -1; + return -1; bufsize = encode_nals(buf, bufsize, nal, nnal); if(bufsize < 0) - return -1; + return -1; /* FIXME: dts */ x4->out_pic.pts = pic_out.i_pts; @@ -117,7 +117,7 @@ X264Context *x4 = avctx->priv_data; if(x4->enc) - x264_encoder_close(x4->enc); + x264_encoder_close(x4->enc); return 0; }